Energy Conversion and Management Strategies for Electro-Hydraulic Hybrid Systems: A Review

Abstract

The electro-hydraulic hybrid system has emerged as a critical technology in new energy vehicles, owing to the remarkable power density and efficient energy regeneration capabilities of hydraulic technology, coupled with the high energy density of electric power. This system effectively enhances vehicle range and battery life. We developed an energy management strategy (EMS) for the electro-hydraulic hybrid system (EHHS) to ensure smooth energy conversion, while ensuring the full utilization of electrical and hydraulic energy within a reasonable and efficient range. To enhance the system’s overall performance, it is imperative to address pivotal technologies, including power coupling and energy management. In this research, the structure of an electro-hydraulic hybrid vehicle (EHHV) is classified, compared and discussed. The application of existing EHHVs is studied. Subsequently, an analysis and summary are conducted on the current status and development trends of EMSs and collaborative operation control strategies (COCSs), and a novel mechanical-electro-hydraulic power-coupled system (MEHPCS) is put forward that successfully converts mechanical, electrical, and hydraulic energy in performance. Simultaneously, other applications of the system are forecasted. Finally, some suggestions for the electro-hydraulic hybrid systems’ future development are made. This study can promote the development of sustainable transportation technologies. The system integrates mechanical engineering, control theory, and environmental science, enabling interdisciplinary methodological innovation. In addition, relevant studies provide data support for policy makers by quantifying energy consumption indicators.

1. Introduction

Faced with the energy scarcity and environmental crisis caused by the continuous growth of the global population, carbon emissions have become the world’s primary challenge [1,2]. Among them, the carbon emissions caused by fuel vehicles account for 7.3% of the total global emissions, making them one of the industries with the largest emissions in the world [3,4]. So far, multiple multinational companies have clearly proposed a “carbon reduction” or “carbon neutrality” schedule. With the continuous improvement of energy technology [5], new energy vehicles have emerged as an alternative mode of transportation to replace traditional fuel vehicles [6,7,8]. Customers like pure electric automobiles for their zero emissions and pollution-free features [9]. However, as the biggest bottleneck, the power battery [10] has some fatal weaknesses. The heavy weight of the battery and the declining capacity make it difficult to accurately predict the range, and its service life is short, which easily causes mileage anxiety [11,12]. Considering the above issues, hybrid vehicles are a suitable choice for improving energy efficiency and reducing carbon emissions.

Hybrid electric vehicles are formally defined as automotive systems’ integrating mechanisms, typically integrating an engine with an electric assembly. The most common are mechanical-electro hybrid, mechanical-hydraulic hybrid, and electro-hydraulic hybrid. The corresponding energy storage modes mainly include fluid energy storage (hydraulic accumulator), mechanical energy storage (flywheel), chemical energy storage (battery), and electromagnetic energy storage (supercapacitor) [11].

Table 1. Comparison of energy storage method characteristics [13,14,15,16,17,18,19,20].

Energy Storage Mode/Characteristics	Hydraulic Accumulator	Flywheel	Battery	Supercapacitor
Construction	Simple, compact, and heavy	Simple, large size and weight	More complex, small size, light weight	Simple and compact
Persistence	Durable storage, fully charged and discharged	Not easy to store for a long time, fully charged and discharged	Durable storage, unable to fully charge and discharge	Short charging time, fast charge and discharge, less energy stored
Energy loss	Friction and heat loss	Mechanical friction loss	Energy conversion loss	Energy loss of equivalent resistance
Control difficulty	Flexible and easy	Poor timeliness	Quick and easy	Difficulty
Safety	Stable	Poor, speed too high	Low, environmental pollution	Higher, green
Energy density	Lower	Higher	High	Medium
Power density	High	Medium	Lower	Higher
Efficiency	High instantaneous energy efficiency	High instantaneous energy efficiency	Medium long-term energy efficiency, but low instantaneous energy efficiency	High instantaneous energy efficiency
Lifetime	Long with heat loss	Long with mechanical wear	Longer, can be charged 400–1200 times	Long with unlimited number of charging and discharging cycles
Application	Heavy vehicles	Auxiliary energy storage device for vehicles	Pure electric or hybrid vehicles	Energy storage device for electric vehicles

depicts the characteristics of four energy storage options, whereas Figure 1 shows the energy storage density distribution.

The strengths of hydraulic energy storage include high power density, a simple and compact structure, and the ability to quickly input and output energy in a short time [13]. The manufacturing cost is low and easy to control, especially in the frequently utilized lifting, walking, and braking system of engineering vehicles. The characteristics of flywheel energy storage are similar to hydraulic energy storage, both of which have high power density [14]. But its volume is large and heavy, requiring lightweight design through material improvement. It has a longer lifespan but a mechanical wear risk, making it more suitable for braking-energy-assisted recovery. The battery energy storage offers advantages such as superior energy concentration, enhanced safety features, and eco-friendly characteristics. However, the overcharging and discharging of the current have a substantial influence on the battery capacity, and the cost is high, with limited charging and discharging times [15]. In addition, the energy density of the supercapacitor [16] is small compared to that of a flywheel or a battery. However, while the energy conversion rate remains notably high, the effectiveness of the power transfer during the charging and discharging cycles is significantly influenced by internal resistance factors. Consequently, hybrid vehicles that integrate electric and hydraulic power can have both high power and high energy density. Electro-hydraulic hybrid technology is better suited to frequently changing working conditions and can enhance electro-chemical energy storage durability with the aid of hydraulic power.

In addition, the efficient recovery and release of hydraulic and electric energy storage methods depend on the rational allocation of energy, which needs to be guided by responsive energy management strategies [17,18,19]. Nevertheless, when hybrid vehicles switch between different operating modes, it is essential to develop a collaborative operation control strategy to ensure the smoothness of the switching process. EMSs have been explored in key research on their energy recovery and utilization of electro-hydraulic hybrid vehicles. From rule-based EMSs to optimization-based EMSs, the economic characteristics of the powertrain are improved under vehicle stability conditions, providing significant theoretical references for hybrid vehicles’ EMSs. However, due to the nonlinear structure and complex operating modes of the power system, as well as the strategies’ requirements for online application, developing an efficient and adaptive EMS still faces enormous challenges. With the rapid advancement of hybrid vehicles, the hybrid degree of the electro-hydraulic hybrid powertrain continues to increase. The diversity of working modes and the increasing classification of hybrid systems have led to increasing difficulty in collaborative control strategies. During the mode transition phase, the electro-hydraulic hybrid system involves torque and speed coordination control of components such as motors, accumulators, hydraulic pumps/motors, and clutches. The motor responds faster than the hydraulic pump/motor, and the accumulator offers a higher power density than the battery, enabling swift energy recovery and release. Thus, we can utilize the fast response characteristics of the motor to compensate for the slower power characteristics of the accumulator, and reasonably control the separation and engagement process of the clutch to avoid severe fluctuations in the total output torque caused by torque imbalance during mode switching, which can cause a significant impact and affect the stability of the entire vehicle and component life. In summary, improving the smoothness of the mode-switching process has great practical significance and broad application prospects for the research on hybrid vehicle technology.

Scholars have conducted extensive research on EHHVs. Research has shown that EHHS is mainly applied in commercial transportation, railway systems, and heavy-duty construction equipment. The research and development of EHHVs has gradually engaged the concern of governments, institutional vehicle manufacturers, and universities [20]. They analyzed the different layout forms of hydraulic hybrid power systems and explored their respective characteristics and applications. Due to the specific operating range of urban transportation vehicles, which often require frequent starting and stopping, the EHHS is an ideal choice for urban vehicles. Sun et al. proposed a new type of electro-hydraulic hybrid system that could be applied to passenger cars, express cars, and trucks in cities [21]. Wu et al. [22] applied the EHHS to medium-sized trucks. There was also a fully hydraulic hybrid system applied to diesel rail cars and an improved electro-hydraulic drive rail system (EH3). Moreover, construction machinery [23,24] uses hydraulic accumulators to improve energy recovery rates, such as excavators [25], dump trucks [26], loaders [27], and forklifts. This approach utilizes a hydraulic pump to transform the gravitational potential energy from the operational apparatus into hydraulic energy. The generated hydraulic energy is then stored within an accumulator, leading to an enhancement of the efficiency of energy usage.

In addition, the application of electro-hydraulic hybrid power systems in the military sector is becoming increasingly widespread, mainly in the following areas.

(1)

Armored vehicles and main battle tanks.

EHHS can achieve electric mode driving for armored vehicles and main battle tanks, significantly reducing thermal signals and noise and improving concealment. The hydraulic system provides high torque at low speeds, making it suitable for heavy-duty vehicles to maneuver complex terrains such as mud and mountains. Recovering kinetic energy through hydraulic accumulators during braking or deceleration can improve fuel efficiency and extend mission range. For example, the GDLS Hybrid ES3 prototype tank, tested by the US Army, achieves short-term burst acceleration through a hydraulic accumulator.

(2)

Unmanned combat platform.

The hydraulic system has high power density and is suitable for powering small, unmanned platforms (tracked robots) and weapon systems. We realized the lightweight design of unmanned aerial vehicles. Hydraulic actuators can complete actions in milliseconds and are suitable for scenarios such as drone catapult takeoff and rapid opening and closing of weapon doors. The adaptability of hydraulic systems to extreme temperature and vibration environments is superior to that of pure electric systems. For example, the British “Titan” unmanned ground vehicle uses an electro-hydraulic hybrid power to perform mine clearance and material transportation tasks.

(3)

Ships and underwater equipment.

EHHS can reduce mechanical transmission noise. The system, in conjunction with batteries, enables the concealed navigation and silent propulsion of submarines. In the electro-magnetic catapult system of aircraft carriers, EHHS is used to assist in energy management and to balance a high power demand. The hydraulic system is resistant to high pressure and suitable for precise control of deep-sea exploration robotic arms.

(4)

Field energy and logistics equipment.

The hydraulic accumulator is linked with a diesel generator to provide instantaneous high-power output (radar or laser weapon pulse power supply), which can be applied to mobile generator sets. Hybrid excavators and bulldozers can reduce fuel pressure in battlefield logistics, such as the hybrid engineering vehicle developed by the US military’s ERDC.

The application of EHHS in the military field is still in a rapid development stage. With the improvement of energy efficiency and reliability, it is expected to become one of the core technologies of the next generation of high-mobility equipment.

This research introduces the structural form and the application scenarios of EHHVs, summarizes the application of existing EMSs and COCSs in power transmission systems, looks forward to the development prospects of EHHs, and sorts out the bottlenecks it faces. The rest of this paper has the following structure. The structural principles of the EHHS are described in Section 2. Section 3 offers a comprehensive summary of the present research landscape concerning EHHVs. Section 4 outlines the energy management strategies for hybrid vehicles and the collaborative operation strategies of different hybrid vehicles are explored in Section 5. The principle and modes of MEHPCS are introduced in Section 6. Section 7 explores the possible development directions and challenges of EHHSs. The thesis is summarized in Section 8.

Drawing from previous analysis, this research combines electro-hydraulic power-coupled systems with energy management to investigate control frameworks for electro-hydraulic hybrid systems. The system strives to optimize energy efficiency without compromising vehicular stability, while proposing five innovative solutions to advance modern electro-hydraulic automotive technologies.

• The configuration characteristics, coupling mechanism, and coupling method are studied for electro-hydraulic hybrid systems.
• The energy management strategies and collaborative operation control strategies for existing hybrid vehicles are explored.
• A mechanical-electro-hydraulic power-coupled vehicle is proposed to introduce the new application of electro-hydraulic hybrid systems in hybrid vehicles.
• An introduction to the applications of EHHSs in other fields, such as wave converters, vehicle suspensions, and wind power generation, is presented.
• Some innovative suggestions are put forward for energy management strategies and cooperative operation control strategies for electro-hydraulic hybrid systems.

2. Electro-Hydraulic Power-Coupled System Configuration

In dynamic power-coupled system design, research has been undertaken by scholars. Liu [2] proposed a series hydraulic hybrid electric vehicle. Through the application of hydraulic accumulators, a mechanical-hydraulic integration was established, enhancing power and energy recovery efficiency while addressing the limitations of conventional vehicle systems. Zhu et al. [28] proposed adaptive energy management strategies and optimal energy distribution strategies for battery-supercapacitors in plug-in hybrid vehicles, achieving mechanical-electric coupling in hybrid vehicles. According to the structural characteristics of mechanical-electric and mechanical-hydraulic power-coupled systems, electro-hydraulic hybrid vehicles are equipped with accumulators and batteries. Based on the energy-transfer path and system connection, power-coupled systems can be classified as series, parallel, and series–parallel. Different forms of hybrid power are clearly distinguished in terms of structure, and there are also differences in the electro-hydraulic coupling mechanism. This section mainly introduces the topology and coupling methods. The empirical formulas for electro-hydraulic power-coupled vehicles are shown below.

(1)

Vehicle model

By applying the vehicle’s kinematic equations, the necessary driving power can be mathematically derived through the following computational procedure. The vehicle power demand in Equation (1) needs to overcome rolling resistance, slope resistance, acceleration resistance, and wind resistance.

$$P_{\mathrm{vehicle}}={\displaystyle \frac{u}{3600{\eta }_T}}\left(mgf+mgi+{\displaystyle \frac{C_{D}A{u}^2}{21.15}}+\delta m{\displaystyle \frac{du}{dt}}\right)$$

(1)

P_vehicle is the vehicle power demand. η_T is the driveline efficiency. u is the vehicle speed, g is gravitational acceleration, i is the gradient, f is the rolling resistance coefficient, C_D is the air resistance coefficient, A is the windward area, δ is the mass rotating coefficient, m is the vehicle mass, and du/dt is the acceleration.

(2)

Motor model

In the EHHS, the motor plays an important role as a conversion device for electrical and mechanical energy. Electric motors can be used as motors to output torque externally, or as generators to charge batteries in reverse. The peak power of the drive motor P_m1 should be configured at 120% of the vehicle’s continuous operational power demand, and the peak power of the control motor P_m2 requires power capacity scaling to 50% of P_m1 during dual-motor cooperative propulsion.

(3)

Battery model

For compliance with driving range criteria, the energy storage system requires optimization to the threshold parameters. Sustained propulsion demand is determined by the continuous power necessary to maintain velocity under pure electric operational state. When driving at a constant speed on a flat road without any slope, the required power of the vehicle is shown in Equation (2).

$$P_{\mathrm{ele}}={\displaystyle \frac{1}{3600{\eta }_T}}\left(mgf{u}_{\mathrm{ele}}+{\displaystyle \frac{C_{D}A{u}_{\mathrm{ele}}^3}{21.15}}\right)$$

(2)

Uniform travel distance S necessitates energy expenditure quantified by the following:

$$W_{\mathrm{ele}}=P_{\mathrm{ele}}t={\displaystyle \frac{S}{u_{\mathrm{ele}}}}{P}_{\mathrm{ele}}$$

(3)

The battery energy output must comply with the energy boundary conditions stipulated in Formula (4):

$$W_{\mathrm{B}}={\displaystyle \frac{U_{\mathrm{B}}{C}_{\mathrm{B}}{\eta }_{\mathrm{k}}}{1000}}$$

(4)

The battery energy should maintain the energy vehicle required. The battery energy must be greater than the steady-state energy demand during constant speed operation, as shown in Equation (5).

$$W_{\mathrm{B}}\ge W_{\mathrm{ele}}$$

(5)

P_ele—steady-state power demand, u_ele—constant speed, W_ele—steady-state energy demand for a constant speed, W_B—the battery energy, U_B-the battery’s average voltage, C_B—the battery capacity, and η_k—the discharge depth.

(4)

Hydraulic pump/motor model

The primary role of the hydraulic pump/motor is to provide supplementary propulsion for the vehicle during uphill travel and rapid acceleration. Therefore, the hydraulic pump/motor’s power capacity should be determined by selecting the higher value between peak climbing power and peak acceleration power, followed by deducting the motor’s output. The peak climbing power is shown in Equation (6).

$$P_2={\displaystyle \frac{u_2}{3600{\eta }_T}}\left(mgf\cos {\alpha }_{\max }+mg\sin {\alpha }_{\max }\cdot +{\displaystyle \frac{C_{D}A{u}_2^2}{21.15}}\right)$$

(6)

The peak climbing power of the vehicle is P₂, and the corresponding speed under this power is u₂. The peak acceleration power can be rewritten as Equation (7), based on the vehicle’s required power calculation, Equation (1).

$$P_3={\displaystyle \frac{1}{3600{\eta }_T}}\left\{{\displaystyle \frac{\delta m{u}_3^2}{3.6dt}}\left[1-{\left({\displaystyle \frac{t-dt}{t}}\right)}^x\right]+mgf{u}_3+{\displaystyle \frac{C_{D}A}{21.15}}{u}_3^3\right\}$$

(7)

The peak power is P₃ in acceleration, u₃ is the corresponding vehicle speed, and P_p/m is the hydraulic pump/motor’s power. The coefficient x is generally set to 0.5, the iteration interval dt is defined as 0.1 s, and the acceleration time t is quantified in seconds.

$$P_{\mathrm{p}/\mathrm{m}}\ge \max \left\{P_2,P_3\right\}-P_{\mathrm{m}1}$$

(8)

2.1. Coupling Mechanism

The power-coupled system mainly consists of a power source, coupler, controller, actuator, and transmission device. In the current dynamic coupling structure, planetary gear mechanisms are mainly used. The THS system, developed by Toyota, is the earliest power-coupled system to achieve mature mass production, and has developed three generations of products [29,30]. General Motors developed a coupling system for a three-row planetary gear train. This configuration achieved six working modes through the mutual combination and separation of different clutches, which can improve the off-road performance in practical applications, but it was complex and difficult to control. The ISG-type power-coupled system used by Volkswagen, Geely, Chery, and Great Wall Motors is currently the most commonly used coupling system in automotive companies. The coupling form is to connect the engine output shaft and the driving motor’s rotor shaft through a clutch, while shafts of the motor rotor and the transmission input are connected through a clutch.

Scholars have conducted extensive research on the structural characteristics and mathematical models of planetary gear mechanisms. Zhang [31] introduced a novel design approach for parallel dual-row planetary gear systems, employing the lever analogy principle through a unified three-node modeling framework that is applicable to both single-stage and two-stage single-row planetary configurations. Utilizing the principles of bond graph theory, Zhang [32] conducted optimization endeavors concerning the configuration of a hybrid electric vehicle’s dual planetary row power-coupled mechanism. In work by Zhuo [33], an innovative multi-faceted energy management approach was introduced for optimizing the hybrid double planetary transmission system. Yan’s investigation [34] designed an arrangement of a two motor power system within a commercial vehicle by a planetary platoon.

2.2. Coupling Method

The coupling methods of the EHHS are fundamentally categorized into series, parallel, and series–parallel. There are differences in connections, control complexity, and energy flow paths among the three dynamic coupling configurations.

2.2.1. Electro-Hydraulic Series Hybrid Power System

The series hybrid system mainly comprises an engine or motor, hydraulic pump, hydraulic motor, clutch, gearbox, and drive axle, as shown in Figure 2. The engine or motor generates power to drive the rotation of the hydraulic pump, subsequently transferring this power to the drive axle through the hydraulic motor. When driving, the hydraulic accumulator can drive the hydraulic pump/motor separately, or only by the engine or motor. In the braking stage, the energy generated from slowing down can be strategically collected and retained in the accumulator, achieving the effect of energy conservation and emission mitigation. This plan is applicable to small and medium-sized city buses, construction machinery walking devices, and low-speed special vehicles. The overall structural layout is simple and compact, with low control difficulty.

Liu [35] improved the structure of a serial hydraulic vehicle: a hybrid loader featuring a coaxial parallel design with an integrated starter generator (ISG). However, he did not analyze the efficiency of the system, especially the mechanical and volumetric efficiency of the hydraulic pump/motor. Based on the above shortcomings, the study [36] proposed a series model with an EMS based on the fuzzy method. A layered EMS to control the torque of the engine and hydraulic pump was established in the research [37]. The study [38] was conducted on the energy recuperation system of an electric-drive hydrostatic vehicle and proposed three energy regeneration strategies—battery mode, accumulator mode, and battery-accumulator mode—as series model 2.

Sun [39] proposed an electro-hydraulic collaborative system as series model 3 for heavy-duty city buses. By actively charging and auxiliary driving, the engine was maintained in an efficient zone, greatly improving the fuel economy and driving performance. Aiming at the shortage of regenerative braking in traditional electric vehicles, Ramakrishnan [40] proposed an energy-efficient electro-hydraulic series collaborative system. An evolutionary strategy algorithm was utilized for multi-goal optimization, with the goal of maximizing energy efficiency.

In summary, the series hybrid system regulates the hydraulic pump/motor’s flow rate, providing ease in managing the vehicle’s frequent starting and stopping conditions. Additionally, it demonstrates notable efficacy in recuperating and utilizing braking energy efficiently. However, due to the multiple transformations of electric/mechanical or hydraulic energy to mechanical energy within the system, the overall efficiency is relatively low.

2.2.2. Electro-Hydraulic Parallel Hybrid Power System

Figure 3 depicts the configuration of the electro-hydraulic parallel hybrid system. The system has two energy flows: one is mechanical power, starting from the motor, and another is converted into mechanical power through the hydraulic pump/motor mechanism. Through the power coupling mechanism, the two are combined and delivered to the drive axle. This system facilitates diverse working modes, such as pure electric driving, pure hydraulic driving, electro-hydraulic driving, and hydraulic or electric regenerative braking. Liu [24] designed an electro-hydrostatic hybrid system. The system consisted of a variable pump, a pump/motor, and a flow modulation valve with a proportional control and accumulator. Ye [41] adopted a rear parallel hydraulic hybrid power system. The results of actual vehicle tests indicated that hydraulic hybrid vehicles reached a 30% energy utilization efficiency, improved their power performance by 16.2%, and decreased fuel consumption per 100 km by 36.8%. However, it could not realize infinitely variable speed and adjust the optimal working area of the pump/motor through speed change.

Li [42] developed an innovative electro-hydraulic hybrid transmission system for vehicles. To mitigate the initial current of the traction motor in rail engineering vehicles, Li [43,44] introduced a technique for managing the motor startup current based on the reverse drive with the hydraulic pump/motor. Results showed that the electric power and energy-saving effect were obvious in the coupling mode of pump inlet flow, but there was a large impact on the system and a higher cost. Yang [45] designed an mechanical-electro-hydraulic power coupler that integrated the motor and plunger pump, greatly reducing the structural size, but the design difficulty also increased. Its input and output are coaxial, and there is only one output—only a fixed transmission ratio output mechanical power—which could not achieve stepless transmission. To achieve continuously variable speed in a parallel hybrid power system, Li [46] proposed a power coupling system design scheme based on planetary gear mechanisms. By incorporating a control motor, the system facilitated the power modulation, together with the driving motor and hydraulic pump/motor.

According to the above analysis, the hydraulic system of the electro-hydraulic parallel hybrid power system does not work full-time and requires less power for the hydraulic system, which has higher adaptability compared to the series form. The transmission route is mainly in the form of mechanical energy, with minimal energy loss. On the premise of meeting power and stability, it can adjust the working area of the motor/engine to achieve energy-saving efficiency. This is widely used in heavy-duty trucks, military trucks, and other vehicles in medium to high-speed mode. However, the future research direction is based on how to ensure that the working state of the motor does not undergo sudden changes and improve energy conversion efficiency under low-speed and frequent start-stop conditions.

2.2.3. Electro-Hydraulic Series-Parallel Hybrid Power System

The electro-hydraulic series-parallel hybrid power system integrates series and parallel hybrid power systems, with a more efficient energy recovery efficiency and utilization rate. The system structure generally consists of a power-coupling mechanism, a hydraulic system, and an electric power system. The motor establishes a connection with the hydraulic pump/motor and power coupling mechanism through a gear mechanism. The power coupling mechanism is joined to the hydraulic pump/motor through a clutch. During the working process, the system can provide multiple power transmission routes in series, parallel, and hybrid, and cooperate with multiple working modes (Figure 4).

Wang [47] identified a new variable pump/motor hybrid hydraulic garbage truck structure that improved fuel economy by 37.2% compared to conventional garbage trucks. Nie [48] created a dual-planet hydraulic hybrid system on the basis of the driving conditions and configuration characteristics of urban public transportation. Sun [49] presented a transient optimal EMS for integrated transmission efficiency, based on the efficiency performance for the hydraulic pump/motor.

Yu [50] presented an energy-conservation scheme for a new hydraulic hybrid excavator and designed an electro-hydraulic stepless transmission system that drove the main pump. Zhang [51] developed an oil-electric hydraulic three-way hybrid system for hybrid heavy-duty vehicles. Li [52] configured a new multi-source hybrid truck system. The engine, motor, and hydraulic pump/motor were designed in a hybrid manner through a planetary gear mechanism, utilizing a continuously variable transmission to attain optimal synchronization.

In summary, the electro-hydraulic series-parallel hybrid power system has shown astonishing advantages for energy distribution and recycling. However, because of its complex structure and difficult control strategy, it requires a lot of data transmission and calculation, which increases the calculation burden of the controller. Therefore, the system is rarely used in real vehicles at present, and more advanced energy management strategies remain to be investigated.

3. Electro-Hydraulic Hybrid Vehicles

From the current development status of hybrid vehicles, hydraulic accumulators with high power density and batteries with high energy density have been widely used in vehicles with electro-hydraulic hybrid systems. After browsing a large number of studies, we know that the electro-hydraulic hybrid system is extensively applied in commercial vehicles, rail vehicles, and engineering machinery. This chapter mainly introduces typical application cases of EHHSs in academic and business circles in real vehicles.

Table 2. Application of electro-hydraulic hybrid vehicles.

No.	Author	Vehicle Type	Energy Sources	Reference
1	Yang, J.	Truck	Battery-accumulator	[53]
2	Hong, J.	Truck	Battery-accumulator	[54]
3	Zhu, Z.	Agrimotor	Battery-pump/motor-engine	[28]
4	Liao, J.	Suburban utility vehicle	Battery-accumulator	[55]
5	Sun, H.	Heavy duty vehicle	Engine-accumulator	[56]
6	Bravo, R.	Heavy duty vehicle	Engine-accumulator	[57]
7	Xia, L.	Excavator	Battery-accumulator	[26]
8	Zhou, S.	Truck	Engine-accumulator	[58]
9	Hao, Y.	Excavator	Battery-accumulator-engine	[59]
10	Niu, G.	Bus	Battery-accumulator	[60]
11	Chen, G.	Bus	Battery-accumulator	[61]
12	Liu, H.	Bus	Battery-accumulator	[62]
13	Liu, H.	Rail vehicle	Battery-accumulator	[24]
14	Liu, H.	Rail vehicle	Battery-accumulator	[63]

lists the typical applications of the electro-hydraulic hybrid system in various vehicle models.

3.1. Commercial Vehicles and Rail Vehicles

More applications of EHHVs in commercial vehicles can be found in reference [53,54,55,60]. The specific application of the electro-hydraulic power-coupled system in trucks and rail vehicles is shown in Figure 5. Yang [45,53,54] proposed a new mechanical-electro-hydraulic coupling electric vehicle (MEH-PCEV), which integrated the motor, accumulator, and hydraulic pump/motor. In the starting stage of the vehicle, the hydraulic power was used to decrease the peak torque at the beginning of the motor. During acceleration or climbing conditions, the accumulator assisted the motor in increasing the output torque. When the load was small and acceleration conditions were not required, only an electric motor provided power. Under braking conditions, regenerative braking energy could be preferentially transformed into hydraulic energy and stored in the accumulator, while the remaining braking energy was stored in the battery through motor braking.

Liao [55] designed an electro-hydraulic hybrid four-wheel driving SUV, using a hydraulic auxiliary power system to address the significant power fluctuations generated by pure electric vehicles during driving and braking. The system included front axle motors, rear axle motors, hydraulic pumps/motors, and high/low pressure accumulators, which could achieve various operating modes. Niu [60] introduced an electro-hydraulic hybrid parallel (EH2) system for urban public transportation. The front axle was driven by a separate motor, and the rear axle was driven by an accumulator and hydraulic motor. Its operating modes mainly included cruise mode, regeneration mode, launch mode, and acceleration assist mode.

Southwest Jiaotong University has conducted many detailed studies on the application of electro-hydraulic hybrid systems in rail vehicles. Liu [24,62,63] proposed an electric hydrostatic hybrid system (EH3) to solve problems such as the poor downhill stability and low energy efficiency of vehicles. The EH3 system consisted of a motor, hydraulic pump, hydraulic pump/motor, high-pressure accumulator, low-pressure accumulator, and several valve groups. In starting mode, hydraulic power was prioritized to eliminate the instantaneous high current impact. The motor and accumulator released energy simultaneously in acceleration mode, eliminating the high current discharge of the battery and extending the range. Hydraulic braking energy recovery could be carried out to improve energy utilization efficiency. If regenerative braking was not suitable, the vehicle entered hydraulic non-friction braking mode as a means to mitigate wear from friction braking.

Commercial vehicles, especially heavy trucks and urban buses, have increasingly strict requirements for fuel economy and emissions. The EHHS significantly reduces fuel consumption and carbon emissions by recovering braking energy and optimizing power distribution. The new generation system reduces space occupation through high-voltage and electro-hydraulic integrated modules (pump–motor integration). We dynamically managed hydraulic and electric power sources through model predictive and real-time optimization algorithms to adapt to complex working conditions, such as urban start-stop and ramp driving. Through the electro-hydraulic architecture, the hydraulic system serves as a “power buffer” and works in conjunction with batteries/fuel cells to solve the instantaneous high-power demand of pure electric commercial vehicles.

The braking energy of rail vehicles is large and frequent, and traditional resistance braking wastes a lot. The EHHS quickly stores/releases energy through hydraulic accumulators, making it more suitable for high-frequency and high-power scenarios than batteries. In contactless railway sections, such as suburban railways and branch lines, the electro-hydraulic hybrid system can replace diesel engines and provide a clean solution. Compared to pure electric driving mode, hydraulic systems are more durable in harsh environments such as humidity and dust, making them suitable for long-term operation on rails.

3.2. Construction Machinery

The EHHS is widely used in construction machinery, mainly in vehicles with hydraulic power such as heavy-duty trucks and excavators [64], even in agricultural machinery [59,65]. Sun [56] proposed a parallel hybrid electric vehicle, composed of a hydraulic/electric collaborative system. The system was applied to heavy-duty vehicles. Focusing on the traits of repeated starting and stopping, along with substantial braking energy, a more effective energy storage method was found. The hydraulic pump/motor was linked to the motor through a clutch, and the power transmitted by the engine through the transmission and the motor was coupled at the torque coupler and output to the drive axle. However, there were certain problems with this structure. In hydraulic propulsion mode, the hydraulic motor drove the motor to idle, which increased the system’s rotational inertia and resistance, resulting in a decrease in power-transmission efficiency. Heavy vehicles generate a large amount of braking energy when braking downhill. To recycle this energy more efficiently, Bravo [57] designed a hydraulic-pneumatic hybrid system. The demand vehicles that this system adapted to were heavy vehicles that needed to change roads and urban areas. The engine and the hydrostatic system were arranged in parallel, and the hydraulic system was in parallel with the pneumatic system. Compressed air and hydraulic systems assisted the engine in driving, and regenerative braking energy was stored in hydraulic accumulators and air reservoirs. But the core of this method was the distribution of hydraulic energy and compressed air energy, with the disadvantage of difficult air sealing.

Ge [66,67] proposed an electro-hydraulic excavator, featuring independent metering input and metering output systems. A 6 t hydraulic excavator test bench was built in this paper. Xia [26] proposed an integrated system for driving and energy recovery, utilizing a three-chamber hydraulic cylinder to address potential energy loss during the excavator arm’s descent. Yu [50] proposed an electro-hydraulic continuously variable power transmission system that drove the boom to rise and recover the braking energy when falling. It mainly consisted of an engine, electric motor/generator, inverter, battery, hydraulic pump, and planetary gear mechanism. Hydraulic motors and generators were used to recover regenerative potential energy. The system achieved an efficient combination of engine and electric motor/generator through a planetary gear mechanism for infinitely variable speed. However, this study still had limitations, as EHCVP was unable to control the engine for a more efficient operation when hydraulic excavators operated under very high loads and moderate speeds.

Zhu [28] designed a mechanical-electronic-hydraulic power transmission system that combined electro-mechanical and hydraulic composite transmission. Aligned with the specific operational demands of the tractor, four driving modes could be achieved: pure electric propulsion, pure engine propulsion, torque-coupled propulsion, and speed-coupled propulsion. Results from the simulation showed that the speed adjustment and efficiency features maintained good agreement across the tractor’s speed range.

Compared to traditional hydraulic systems, EHHSs in construction machinery (excavators, loaders, forklifts) significantly improve energy efficiency and reduce emissions through energy recovery, power optimization, and intelligent control. Under frequent start-stop and rotation-braking conditions, construction machinery uses hydraulic accumulators to store braking energy for auxiliary power output. Electric hydraulic pumps can replace engines’ direct drive pumps to achieve precise flow control. In the future, digital hydraulic and AI control will become mainstream. The degree of system lightweighting and intelligence has been greatly improved (Figure 6).

4. Research on Energy Management Strategies

The key technology of hybrid vehicles cannot be separated from EMSs [68,69]. EHHS has two or more power sources, corresponding to a variety of operating modes. Accordingly, multiple power sources also bring complexity to the power transmission system, which requires certain EMSs to distribute energy flow and achieve mode switching. The primary role of EMSs is to manage the vehicle’s optimal control strategy across various states while meeting the driver’s requirements. It mainly solves the following problems:

(1)

The design of driving and braking EMSs for each operating mode, while ensuring stability.

(2)

From the current working mode to the next moment of the working mode, completing the power source and actuator speed and torque control strategy.

(3)

The design of EMSs, to improve the efficiency range of the motor or engine or hydraulic pump/motor in each operating mode.

(4)

To ensure smooth operation during deceleration, a regenerative braking approach should be implemented to minimize abrupt transitions in both the electric motor and battery system. The strategy should focus on gradual energy recovery while maintaining system stability.

In this research, EMSs for hybrid electric vehicles are mainly divided into two categories: rule-based EMS [70,71,72] and optimization EMS [73,74,75]. The specific EMSs classification scheme is shown in Figure 7.

4.1. Rule-Based EMS

In EMSs, rule-based strategies are relatively simpler methods. The rule-based EMS is formulated to optimize performance and relies on engineering experience, analysis of related components, efficiency characteristics, and mathematical models of hybrid power systems. It is one of the most widely used strategy methods. This type of EMS typically includes both deterministic rules and fuzzy rule EMSs.

4.1.1. Based on Deterministic Rules

The EMSs based on deterministic rules can select logical threshold values, according to vehicle operating conditions and existing experience. By determining the vehicle’s driving information parameters, the vehicle enters the given operating mode. The EMSs based on deterministic rules are displayed in Figure 8.

Sun [56] proposed a parallel hybrid vehicle structure composed of a hydraulic/electric collaborative system: the development of a control strategy that combined logical threshold methods and key parameter optimization algorithms to address the energy waste caused by frequent starting and stopping and extensive braking of heavy vehicles. The proposed EMS had great potential for fuel economy and maintained SOC bias within the high efficiency range. Meng [76] proposed a rule-based EMS, designed to enhance braking energy recuperation efficiency while addressing the challenges posed by high-speed energy recovery on the hydraulic system in a unique mechanical-electro-hydraulic hybrid electric vehicle. Yang [45] designed a rule-based dynamic optimal EMS to control energy allocation and mode switching in real time. The MEH-PCEV demonstrated a reduction in battery consumption of 14.418% and 21.174% in NEDC and UDDS, respectively. Moreover, the time to reach the starting speed threshold increased, and the electrical peak torque significantly decreased. The above researchers have all proposed an electro-hydraulic hybrid power system and adopted a rule-based EMS to improve energy utilization efficiency.

In energy-saving characteristics research, Gong [77] designed a new electro-hydraulic energy-saving system on a 23 t hydraulic excavator, implementing real-time control using a parameterized rule-based strategy. Experimental research showed that the system saved about 17.6% energy, but he did not consider the comprehensive optimization of control strategies and components. Zhang [78] developed a rule-based EMS and proposed a parallel electro-hydraulic hybrid system. The lifespan of electric wheel loaders was extended by integrating hydraulic power systems to supply and recover energy during start-up and braking operations. This strategy enabled the parallel electro-hydraulic hybrid system to reduce fully equivalent battery cycles, boosting battery life by 15.64%, in contrast to purely electric configurations. When studying battery life, the impact of the hydraulic system on its lifespan was not taken into account. He [38] designed an acceleration and braking strategy for a battery-operated hydrostatic propulsion vehicle, based on logical threshold values. To lower the energy inefficiency of hydrostatic elements, Liu [24,63] proposed an EH3 power system to enhance the energy efficiency and downhill speed stability of electric rail vehicles. EH3 adopted a hydrostatic system, and the energy recovery efficiency of a traction motor can reach 40%, which ensures that the traction motor can recover energy efficiently and stably. Faced with minimal elevation changes, the traction motor demonstrated reduced energy harvesting capacity, yet sustained 20% operational efficiency. To solve the speed regulation problem of downhill friction braking, a control method was proposed, which only needs to control the displacement and opening of a variable displacement pump/motor. Additionally, to better solve the challenges associated with inadequate energy utilization and the significant power demand spikes in traction motor systems, hydraulic regenerative braking (HRB) was used instead of friction braking (FB). The hydraulic energy recovery rate reached 50% with a hydrostatic system. The proposed hydraulic energy coupling idea provides a basis for the further study of electro-hydraulic hybrid power systems. However, RB-EMS is widely used in this paper. The mode-switching conditions of this strategy are fixed. If the external environment changes, whether the mode-switching conditions can be tuned according to the changes in the external environment remains to be studied.

Based on the aforementioned electro-hydraulic hybrid model, Chen [61] developed a rule-based dynamic optimal EMS to solve high peak torque and short battery life during starting and acceleration for urban electric buses. Under NEDC conditions, the peak torque decreases by 36.4% and the battery power consumed decreases by 33.98%. Under actual conditions, the maximum torque output experienced a reduction of 28.9%, accompanied by a 32.3% decline in battery power consumption. This control strategy can effectively reduce the peak torque of the motor, indirectly improve the service life of the battery, and provide a reference for electro-hydraulic hybrid power systems. Sun [39] proposed a hydraulic/electrical collaborative system (HESS) to address the limitations of one energy storage system for heavy-duty hybrid vehicles. An EMS specifically designed for collaborative systems was created to manage power allocation between multiple energy sources.

Although numerous studies have adopted rule-based EMSs and achieved good results, for complex electro-hydraulic hybrid systems in special scenarios, the parallel distribution generated by multiple parameter combinations makes it difficult to accurately select logical threshold values. Secondly, it is difficult to accurately delineate the boundaries between working modes, resulting in the inability to switch quickly between multiple power sources in a short time, and the inability to ensure optimal energy efficiency. In addition, it is usually necessary to use other optimization methods to develop more efficient and energy-saving control logic, to achieve better power and economy when setting logical threshold values. Consequently, EMSs based on deterministic rules have great optimization space.

4.1.2. Based on Fuzzy Rules

The EMS based on fuzzy rules is another manifestation in rule-based EMS, which can be seen as the optimization based on deterministic rules. Compared with the deterministic rule-based EMS, the state variables are divided into fuzzy sets. This method calculates the membership of each rule based on the fuzzy degree and then makes corresponding decisions.

Yang [53] adopted a reverse approach, involving fuzzy logic optimization EMS (FLO-EMS) with multi-parameter objectives as inputs, to address the phenomenon of abnormal output torque in RB-EMS. A fuzzy controller included extreme torque, accumulator pressure, and vehicle speed as variables, and optimized the logic threshold strategy through an inverse multi-parameter collaborative control method. Fuzzy control can be used to optimize the torque mutation of the motor output torque, while optimizing the logic threshold. However, fuzzy control requires a large number of empirical parameters, which are difficult to grasp in real time. To more accurately reflect the relationship between vehicle operation mode and fuzzy control, Liu [62] designed an EMS combining driving pattern recognition (DPR) and fuzzy logic rules (FLR) with the EH3 powertrain as a reference. Active regulation and bidirectional transmission of electric motors and variable displacement pumps/motors were achieved. Based on RB-EMS, this paper combined driving modes and speed in an actual driving scenario with the learning vector quantization neural network and introduced the fuzzy controller to process multiple variables. The study revealed that implementing DPR and FLR mechanisms in EMS allows for optimized power-flow management in hybrid configurations, leading to notable enhancements in battery durability and vehicle range. Xiong [79] demonstrated a power system that can control mode switching and instantaneous power distribution between series and parallel mode. Under the buses’ driving conditions, the theoretical energy consumption by fuzzy logic as the mode decision criterion exhibited a reduction of 30.3% compared to conventional buses.

To solve the optimal power distribution problem between electric motors and hydraulic power, the following researchers have conducted extensive research between powertrain and fuzzy control. Silva [80] designed a dual HESS electric powertrain and fuzzy control to seek the optimal design variables related to the maximum driving range and battery life. In the same year, Miranda [81] proposed a fuzzy logic control strategy to achieve the optimal power allocation between motors while maintaining EV performance and reducing energy consumption. By maximizing the battery state of charge (SOC), the mass of the motor and battery were minimized, and the driving range was extended. Eckert [82] applied optimal fuzzy logic control to combine hydraulic hybrid electric vehicles. The validation results obtained under the FTP-72 standard driving cycle showed that HCCONOx reduced emissions by 13.07% and saved fuel by 35.67%.

Li [83] proposed an adaptive real-time energy management control strategy for commuting hybrid electric vehicles. The simulation results revealed that, contrasted with rule-based strategy, traditional ECMS, linear SOC reference A-ECMS, A-ECMS that PSO optimized EF, and SOC-fixed PI feedback-regulated A-ECMS, the average fuel consumption was reduced by 22.98%, 10.26%, 6.52%, 2.33%, and 5.91%, respectively. This strategy incorporated the reference SOC of the fuzzy inference system (FIS) and the adaptive equivalent consumption minimization strategy (A-ECMS). This approach dynamically modulates the strategy’s adaptive coefficient by incorporating real-time road data alongside continuous optimization based on the power storage state. Shi [84] proposed a new fuzzy adaptive method for the optimal strategy of PMP. A fuzzy adaptive algorithm was formulated to compute the collaborative state by employing the PMP strategy, incorporating both the anticipated average power and the real battery SOC. The algorithm optimizes the collaborative problem brought by basic fuzzy control through fuzzy adaptive control, which can solve the threshold selection problem under different working conditions (Figure 9).

Fuzzy rules have strong adaptability and robustness and can be easily adjusted according to different needs. Although this method can achieve adaptive suboptimal power allocation, a good fuzzy EMS requires expert knowledge and finding reasonable experts to formulate it is difficult. In addition, the robustness of fuzzy EMS is still poor, and corresponding adjustments need to be made according to different operating conditions. However, the primary limitation stems from its heavy reliance on the designer’s expertise and judgment, which limits the performance optimization of the strategy and the ability of different conditions. Therefore, although the rule-based EMS occupies a place of hybrid electric vehicles, it cannot maximize the fuel-saving potential of hybrid electric vehicles.

4.2. EMS Based on Optimization

The rule-based EMS draws from existing research experience, and the selection of parameters, such as logical threshold values and membership functions, has strong subjectivity. According to Section 4.1, the power consumption or fuel consumption of vehicles decreased by about 5–30% in different strategies proposed in papers. The recovery rate of vehicle braking energy can reach 40–50% [24,63]. The peak torque of the motor decreased by 36.4% [61]. The emission performance decreased by 13.07% [82]. Although it has obvious advantages in terms of vehicle economy and emissions, real-time operation, processing speed, etc., it cannot ensure the vehicle’s optimal overall performance.

The optimization EMS mainly focuses on real-time optimization and global optimization. This control strategy seeks to optimize powertrain performance for enhanced overall efficiency. Furthermore, alongside the accelerated advancement of 5G technology and intelligent systems, intelligent EMSs are increasingly being applied to vehicle EMSs.

4.2.1. Real-Time Optimization

The goal of the real-time optimization strategy is to allocate the optimal power based on the principle of minimizing energy consumption within each step size. It mainly includes equivalent consumption minimum control (ECMC), robust control, model predictive control (MPC), and decoupling control.

Liao [55] proposed a real-time controllable EMS, based on instantaneous efficiency optimization, to optimize the instantaneous energy consumption of vehicles. Testing outcomes based on CLTC-P revealed that this strategy improved the overall energy consumption by 3.27% compared to rule-based EMS and 2.63% compared to global optimization. Adopting an instantaneous optimization strategy can reduce energy consumption compared to a global optimization strategy. Yu [50] proposed a continuously variable electro-hydraulic power transmission system for hydraulic hybrid excavators. An advanced ECMC was designed to lower energy consumption by computing control instructions for the engine, electric motor/generator, and hydraulic pump. The energy-saving efficiency range was 36.69% to 45.16%. However, if the system operated at high loads and moderate vehicle speeds, even without the ECM, the engine’s operating point may still be in the high-efficiency range. Wu [85] focused on the research on a multi-degree-of-freedom parallel multi-gear hybrid propulsion system for heavy vehicles. After adopting the ECMC, the NEDC cycle demonstrated a 30.14% reduction in the rate of fuel usage.

Due to the difficulty in achieving global optimization of plug-in hybrid systems through rule-based EMS, many researchers have embedded model prediction algorithms into hybrid systems. Liu [86] integrated the driving conditions of short-term prediction models and established random prediction and machine learning. The proposed prediction algorithm was applied to the EMS of plug-in hybrid electric vehicles. The above research does not have real driving cycles to verify the accuracy of the strategy. Luo [87] took the P2 configuration plug-in hybrid system as the research object, combined the model predictive control with dynamic programming, and solved the optimal engine torque sequence in the time domain of vehicle speed prediction with the objective of minimizing engine fuel consumption. Relative to rule-based EMS, the predicted EMS, based on radial basis function neural networks, reduced energy consumption by 13.8% under eight NEDC operating conditions. Yang [88] established prediction models for multi-stage Markov and neural networks based on model prediction methods to predict the operating conditions. Modeling outcomes indicated that the model’s predictive EMS, based on a neural network, could approximate the optimality of a dynamic programming algorithm. Jin [89] established an EMS based on model-predictive control for a dual motor-coupled power system model. This strategy effectively improved the overall vehicle economy and reduced driving costs by 12.6%. Jiang [90] used a linear time-varying model predictive-control algorithm to conduct actual vehicle verification. The results showed that the linearization process of this algorithm was simple, with good real-time performance and a good control effect. The above ECMs use MPC to improve the vehicle economy. The results indicate that the MPC has good economic efficiency.

To adapt vehicles to complex working conditions, multi-objective optimization and neural network optimization have also been introduced into ECM. Yang [91] studied a real multi-objective optimization-guided MPC strategy, according to the characteristics of frequent acceleration and deceleration for passenger cars and complex working conditions. The strategy combined multi-source data fusion technology to build a speed prediction controller and SOC reference generator to coordinate fuel economy, gear shift stability and real-time online optimization control. Zhu [92] proposed a real-time energy management system and optimization strategy for electric vehicles, based on deep short-term and short-term memory neural networks, to make global optimal arrangements when connecting electric vehicles to the power grid during charging and train a learning network using historical load information to solve the historical optimal solution, guiding new real-time optimization. This strategy combined historical task information to guide real-time optimization, but it did not consider the curse of dimensionality and real-time computational efficiency issues (Figure 10).

Real-time EMS can collect data in real time, perform dynamic prediction analysis, and use efficient optimization algorithms to optimize the energy system. It has the ability to make fast decisions, enable precise control, and monitor feedback in real time. Based on the analysis of the above research results, it can be seen that the instantaneous optimization control strategy is not significantly affected by changes in vehicle driving scenarios, has strong real-time performance, and can effectively achieve the instantaneous energy-optimal control of hybrid power systems. However, the control form has potential defects, such as frequent power mode switching, which cannot guarantee global optimization.

4.2.2. Global Optimization

According to optimization theory, the total of the local minimums does not match the global minimum. For EMSs of hybrid vehicles, the optimal fuel economy and emission performance throughout the entire operating process are ideal goals. Global optimization methods generally require a certain amount of prior knowledge, which specifically refers to future global operating conditions in energy management problems. Therefore, the essence of this type of algorithm is offline optimization, which determines the global optimal parameters of the dynamic system, based on the constraints of variables after obtaining a large amount of prior data or parameter sets. The commonly used global optimal methods include dynamic programming (DP), game theory, genetic algorithm, particle swarm optimization (PSO), etc.

Ning [93] used the overall energy regeneration volume as the cost function and the pump displacement and hydraulic coupling transmission ratio as decision variables, and adopted DP to identify the global optimum for both pump displacement and transmission ratio parameters, specifically addressing urban operational environments. However, the paper lacked a real-time strategy for brake energy recovery based on real parallel hydraulic hybrid buses. Guo [94] applied DP to process path information and obtained a set of optimal charge state trajectories for calculating the cost function. The proposed cost function saved 5.9% and 10.8% of fuel consumption. This method applied actual road data to system bench tests, which was more realistic. He [95] proposed the first dynamic coordinated control strategy with the variable backup electric mechanism power for electro-hydraulic composite braking. The deviation between the achieved braking force and the intended braking force was significantly decreased. Wang [96] proposed an EMS for hydraulic hybrid wheel loaders, based on DP algorithms. Zhang [97] incorporated driving pattern identification into dynamic optimization approaches, enabling efficient power distribution among the engine generator, battery pack, and supercapacitor in plug-in hybrid electric vehicles. Two researchers integrated driving pattern recognition into the DP algorithm to achieve a reasonable allocation between different energy sources.

To obtain better global optimization results, scholars borrowed bird swarm features and adopted the PSO algorithm for optimization [98,99]. Chen [100] selected the dynamic PSO algorithm to develop energy management and transmission shift optimization strategies for hybrid vehicles. Results showed that compared to the baseline situation, the improvement in equivalent fuel and energy consumption was 30.75% and 59.55% under the optimization of two variables. Wen [101] innovatively integrated a two-level collaborative parameter optimization method based on PSO and DP algorithms. The optimized driving efficiency was improved by 12.19%, with a maximum increase in operating mileage of 16.3%. Sameh [102] used chaos-improved generalized particle swarm optimization (CIGPSO) to optimize the torque of motors and diesel engines in long-term power management.

Scholars have innovated genetic algorithms to dynamically adjust the optimization parameters in the optimization algorithm over time. Eckert [103] developed a holistic optimization framework for the powertrain and control mechanisms of series electro-hydraulic hybrid vehicles, employing an adaptive weight genetic algorithm with interactive capabilities. As a peak power buffering unit, hydraulic accumulators can effectively reduce battery aging. Hybrid vehicles equipped with hydraulic fuel cell auxiliary power units will be a promising alternative solution. In the same year, team member Silva [80] applied the i-AWGA to the multi-objective optimization of an electric vehicle powertrain. Huang [104] designed a driving state identifier in the power transmission system of a concrete mixer truck, which was composed of a hybrid optimization-based random forest algorithm and a genetic algorithm-optimized differential controller. The system transitions to the most efficient operational mode by analyzing real-time vehicle performance metrics. The results illustrated that the performance of the optimized driving state recognizer was improved by 2.50% compared to before optimization. This experiment should utilize big data technology, as more real-world driving information can improve the effectiveness of energy management strategies. Vignesh [105] utilized neural fuzzy adaptive ECMS for intelligent energy management, achieving better fuel economy and lower emissions for plug-in parallel hybrid vehicles. For the self-developed real-world driving cycle, it also achieved a terminal SOC of 27.53% and a fuel economy of 33.37 km/L (Figure 11).

Unlike real-time optimization, global optimization searches the whole region; it can solve the global optimal solution and avoid falling into the local optimal trap. Many global optimization algorithms (genetic algorithms, particle swarm algorithms) simulate biological behaviors in nature, such as evolution, foraging, etc., and this heuristic design makes the algorithm flexible to deal with complex problems. Nevertheless, such algorithms require strong data processing and computing capabilities, as well as the ability to predict complete driving cycle information. The method is capable of obtaining theoretically optimal solutions, but the computational burden is too heavy for on-board computers. Therefore, both real-time optimization and global optimization strategies can be further improved through intelligent optimization algorithms. The method has randomness and uncertainty, which can lead to different results from run to run, increasing the unpredictability of the results. Although global optimization techniques are capable of identifying the most optimal solution, during the final stages of the optimization process, as the solution approaches the global optimum, the rate of convergence may decrease significantly, leading to reduced computational efficiency.

4.2.3. Intelligent Optimization

With the development of high-speed information such as V2V and V2I, vehicles can access the various surrounding information around traffic scenarios in real time. In addition, with the rapid development of technologies such as interconnectivity and autonomous driving [106], it has become possible to optimize the powertrain of hybrid vehicles. In this research, intelligent optimization mainly includes three categories: reinforcement learning (RL), pattern recognition optimization (PRO), and machine learning (ML).

Wang [107] studied 13 popular deep reinforcement learning (DRL) EMSs, providing guidance for developing a unified performance evaluation benchmark, but he did not specify the weight allocation of the evaluation indicators, which may lead to biased conclusions in different scenarios. Zhang [108] combined Q-learning with deep neural networks to construct a dual deep Q-network (DDQN) EMS to solve traditional control strategies and RL problems. Zhang [109] adopted an intelligent energy management architecture based on RL to allocate the power of hybrid engineering machinery. Although the dual network structure can alleviate the problem of overestimation of Q-values, it increases computational complexity and may limit real-time performance. Wang [110] introduced a strategy to minimize energy consumption in hybrid electric vehicles through multi-agent reinforcement learning, ensuring synchronized powertrain control and the following behavior. Sun [111] proposed an EMS based on deep learning and improved MPC, considering the reasonable allocation of power requirements for plug-in hybrid electric vehicles. In many reinforcement learning EMSs, reward functions are typically designed based on subjective judgment and practical experience, making it difficult to describe the intentions of experts objectively. In this regard, Lv [112] proposed an EMS that utilizes inverse reinforcement learning to derive reward function parameters from expert operation patterns, which were subsequently applied to coordinate the actions of engine- and battery-control agents. The above studies have the following problems: the training cost of DRL models is high, it is difficult to adapt to unknown working conditions, and the online learning ability is insufficient. The setting of the reward function is somewhat rough. Digital twin technology and transfer learning can be introduced to improve model generalization.

To achieve better fuel economy and vehicle stability, many scholars have combined RL with optimization EMSs to unlock the energy-saving capabilities of hybrid systems. Dai [113] proposed a layered EMS that combined ECMC and DRL, taking the third-generation Prius hybrid electric vehicle as the research object. Compared with rule-based EMS, this hierarchical strategy could improve fuel economy by 20.83% to 32.66%. Increasing the predictive information of intelligent agents on vehicle speed could further reduce fuel consumption by 5.12%. Zhang [114] proposed a model-based reinforcement learning framework of hybrid engineering vehicles, which avoided cumulative errors between operating cycles and improved long-term learning stability. This strategy introduces an environment model, so that the agent no longer relies on repetitive work cycles of frequent actual interactions. It can accept virtual input to achieve a more efficient and cost-effective learning process.

Due to the complexity and diversity of driving conditions, there are so many uncertainties that are greatly influenced by the driver’s driving style in terms of fuel economy and power. Hence, to improve the energy efficiency and timeliness of hybrid vehicles, many researchers have combined driving style and EMSs. Guo [115] combined driving style with ECMC, employing a hybrid genetic algorithm called PSO-GA to optimize the connection between driving style and the equivalent factor. This led to the development of an adaptable control strategy for optimizing driving techniques. However, the paper did not consider the instantaneous effects at the edge of dynamic operating conditions. Tian [116] identified the driver’s driving style using the offline part. The online section incorporated the driver’s driving style into the ECMC. The proposed strategy increased fuel economy by 9.54% and 7.03%. Offline classification relies on historical data, and sudden changes in driver style may lead to strategy failure. Zheng [117] proposed a driving condition recognition method, based on a principal component analysis-learning vector quantization (PCA-LVQ) neural network intelligent algorithm. The adoption of data-driven methods avoided reliance on expert experience. Qiu [118] proposed a braking energy recovery method that considers the coupling effect of working conditions and driving style by analyzing the impact of working conditions and driving style on braking energy recovery. The above strategies are mostly based on static working conditions and do not fully consider the instantaneous working conditions of implementing environmental perception.

To balance the advantages and disadvantages of the above intelligent optimization algorithms, there are also EMSs that combine multi-objective optimization, deep prediction, machine learning, and reinforcement learning. Zhang [119] proposed an uncertain EMS for hybrid electric vehicles. A hybrid speed predictor, merging convolutional neural networks and short-term memory neural networks, was employed to analyze temporal features and elucidate the patterns of speed fluctuations. The multi-objective optimization problem based on learning is a key issue in EMSs of multi-power hybrid systems and has received widespread attention in recent years. Ruan [120] proposed a hierarchical EMS based on discrete continuous hybrid action to optimize the distribution of dual-motor driving force in battery-power driving and regenerative braking. Huang [121] applied the dual-depth deterministic strategy gradient algorithm framework to the dual mode operation scheme of extended range fuel cell hybrid electric vehicles. Liu [122] established a predictive framework that leverages deep learning techniques combined with multi-source information integration to anticipate short-term operating conditions. Apply this model to the EMS framework based on MPC. The optimized EMS improved fuel economy by 3.32% compared to the traditional MPC. Deng [123] proposed an adaptive equivalent fuel consumption minimum control strategy that adjusted the equivalent factor online, based on changes in operating conditions. The simulation results showed that under the combined working conditions of multiple operating conditions, the fuel economy was improved by 4.18%, and the SOC fluctuation was reduced by 43.26%, demonstrating the superiority of the EMS (Figure 12).

Intelligent optimization has a strong global searching ability and the ability to adapt search strategies and parameters to address different problems and needs. Some intelligent optimization methods also have the capability of parallel processing, which can further improve optimization efficiency. Intelligent optimization methods usually have a clear algorithmic flow and easy-to-implement code structure, which makes them easily integrated into the control system and simulation software of hybrid electric vehicles. Intelligent optimization methods can also be combined with other technologies to form more advanced optimization schemes. By comparing and analyzing the characteristics and performance of different optimization results, the design scheme that best suits the current working conditions and needs can be selected.

The intelligent optimization method has many advantages in hybrid vehicles, but it also has some shortcomings: methods usually require a lot of iterative computation and search operations, which leads to a large amount of computation. Some intelligent optimization methods have high algorithm complexity, such as coding, selection, crossover, and mutation, and other operations in the genetic algorithm need to carry out complex calculations, which increases the difficulty and cost of algorithm implementation. Intelligent optimization methods need to rely on high-quality data for training and validation. If there are problems such as noise, missing data or errors in the data, it may affect the accuracy and reliability of the optimization results.

Although intelligent hybrid vehicles have shown astonishing advantages over ordinary optimization strategies, in terms of research results, their practical application is relatively difficult. Therefore, the study, improvement, and optimization of control algorithms to determine EMSs with strong real-time performance and good control performance have an important engineering application value for advancing electro-hydraulic hybrid system technologies.

4.3. Critical Comparisons

In EMSs of electro-hydraulic hybrid systems (EHHS), rule-based, optimization-based, and intelligent methods each have their own advantages and disadvantages. The bottlenecks in balancing real-time application and practical expansion are as follows:

• Balance of Three EMSs

(1) Rule-based EMS

Rule-based EMS has strong real-time performance and simple logic, is suitable for hardware deployment, and is easy to implement and debug in engineering, but it relies on expert experience, and it is difficult to cover all conditions, which may not be the optimal global solution. It also has poor adaptability to dynamic working conditions, requiring repeated parameter tuning.

(2) Optimization-based EMS

Global optimization: The theoretical optimal solution can be obtained when the driving cycle is known, but global information is required, which is only applicable to offline simulation.

Real time optimization: This balances real-time performance and optimality through rolling optimization, but requires high model accuracy and computational power.

The above calculations have high complexity and require high-performance processors, which may be difficult to meet with millisecond-level control requirements.

(3) Intelligent EMS

The strategy can combine real-time data and predictive capabilities, dynamically optimize energy allocation, and balance economy with strong adaptability of fuel cell life. The optimal strategy for complex working conditions can be learned through data-driven learning. There is no need for precise modeling and it is suitable for nonlinear systems, but the training data demand is large, and the generalization ability depends on the coverage range of the training set. Real-time deployment requires hardware acceleration, which may pose a “black box” risk.

2.

Applicability of real-time application strategies

Simulation research is suitable for using global optimization (DP) or complex intelligent algorithms (deep reinforcement learning) for theoretical validation and benchmark testing. Priority of real-time applications: rule-based > MPC > intelligent algorithms.

Rule-based EMSs have high reliability and meet real-time hard constraints, making them still mainstream for mid- to low-range vehicle models. MPC needs to simplify the model (such as linearization) or shorten the prediction time domain to reduce computational complexity. Intelligent algorithms need to be quantified or pruned to adapt to the embedded hardware, and have gradually been applied in high-end car models, such as the GAC AI platform, achieving “beyond visual range road condition optimization” through deep learning.

3.

The bottleneck of EHHS in real vehicles

The application of EHHS in real vehicles faces technical bottlenecks in hardware, algorithms, costs, and other aspects.

(1) Hardware limitations

In the hydraulic-electrical dynamic coupling system, the high-frequency response of the hydraulic system and the synchronization of motor control require a high bandwidth controller. High precision pressure sensors and high-speed valve components are expensive.

(2) Real-time challenge

Optimization and intelligent methods require high-performance ECUs, but the computing power of automotive grade chips is limited. The hydraulic system requires millisecond-level control, and traditional CAN communication may be delayed.

(3) Reliability and robustness

The difference between actual road conditions (such as slope, load capacity) and simulation may lead to strategy failure. Hydraulic system leaks, battery aging, and other issues require online adaptive strategies to increase software complexity.

(4) Standards and integration

The hydraulic-electrical architecture of EHHS is diverse, making it difficult to develop a universal EMS and it lacks standardization. If integrated with traditional vehicles, it needs to be compatible with existing vehicle electronic architectures, which increases the complexity of the software layer.

In summary, MPC or hybrid strategies (rule + online optimization) are preferred for real-time applications, balancing performance and computational complexity. The areas that need to be broken through in the future are hardware improvement (vehicle specification level computing power), algorithm improvement (online learning under edge computing), system reliability development (hardware in the loop-testing to verify reliability), and data-driven life prediction.

5. Cooperative Operation Control Strategies

Currently, the mode-switching strategy is widely used in mechanical-electrical hybrid vehicles, and there are few references for EHHVs. This section mainly refers to mechanical-electrical hybrid vehicles to describe collaborative operation control strategies. Coordinated control strategies are primarily implemented to manage the interaction between the engine and motor during operational mode transitions, minimizing power output variations, enhancing transition smoothness, and ensuring consistent driving torque delivery [118,124,125,126]. The dynamic coordinated control methods for hybrid vehicles mainly include model prediction [86,87], online engine torque estimation [88,89,90], and torque compensation [127,128].

When the target torque of the engine or motor changes significantly, we should coordinate and control the torque of the motor, engine, and clutch before the engine or motor reaches the target torque, in order to ensure a smoother total output torque of the system and the comfort and power of the mode-switching process [129]. The above issues are also difficulties in dynamic collaborative switching and are a popular research topic for scholars both domestically and internationally.

5.1. Mode-Switching Performance Evaluation Indicators

There is still no unified standard for evaluating the mode-switching performance of hybrid passenger vehicles. We can refer to the requirements of the shifting process [130] to evaluate the mode-switching performance of hybrid passenger vehicles [131].

(1)

Mode-switching time

Mode-switching duration represents the time span from the controller’s command execution to the full establishment of the new operational mode, with its rapidity directly affecting the vehicle’s power system performance. So, the shorter the mode-switching time, the better it is from the perspective of the vehicle’s power. However, the shorter the mode-switching time, the higher the torque change rate, resulting in a decrease in the comfort of the driver and passengers. Thus, the comfort and dynamics of the mode-switching process need to be weighed.

(2)

Impact degree

Impact refers to the vehicle’s change rate in the longitudinal acceleration while it is in motion [132]. The less significant the impact during mode switching, the smaller the torque ripple produced by the hybrid powertrain system, and the smaller the impact on the driver and transmission components. Reducing the impact during mode switching can not only improve driver comfort but also enhance the service life of transmission components.

(3)

Sliding work

Sliding work refers to the energy expended due to the relative motion between the clutch’s friction plates, which is the main source of heat. Most mode-switching processes involve the engagement and disengagement of the clutch. The clutch engagement process includes an empty stroke stage, a sliding wear stage, and a fully engaged stage. In the sliding wear stage, the friction plates at the main and secondary ends of the clutch generate a large amount of heat due to mutual friction, and the friction plates absorb a large amount of heat, causing their surface temperature to rise. The consequence is that the the friction coefficient of plates is affected, leading to increased wear and thereby reducing the service life of the clutch.

5.2. Dynamic Collaborative Control Based on Model Prediction and Online Estimation

This method is based on the low-order vehicle dynamics model to analyze the dynamic response of the system. It employs the optimal control algorithm to calculate the optimized target values of the torque of the engine and the motor. The literature [133] categorized the mode-switching problem as the switching problem of hybrid systems. Through the conversion between different sub-domains of hybrid systems, a method was studied to reduce the impact during mode switching, achieving a seamless shift from pure electric mode to hybrid driving mode. The literature [134] constructed a modal conversion architecture from the pure electric mode to the hybrid power drive and dynamic coupling models of engines and motors under different operation modes. Based on this, a zero-interference mode-switching control method was proposed, which suppressed the amplitude of longitudinal acceleration fluctuation during the pure electric-hybrid power drive mode-switching stage, and effectively improved the vehicle’s operation quality. They are all based on oil–electric hybrid power systems, and, although there are references to electro-hydraulic hybrid power systems, there is still a significant gap. Yang [135] introduced the adaptive rolling horizon control method into the mode-switching strategy of hybrid electric vehicles. A novel adaptive factor was introduced to dynamically adjust the prediction time domain length and control constraints. This factor ensures the ratio between the traction motor’s maximum torque and the system’s required torque is optimally adapted. Zhang [136] designed a torque coordination control strategy based on MPC, with the goal of reducing vehicle impact and clutch slip work. The results showed that electromechanical composite transmission effectively reduced output torque fluctuations and vehicle impact, while ensuring a faster mode-switching response speed, and reduced clutch sliding loss. The use of adaptive factors to dynamically adjust the predicted time domain length, as well as the adoption of torque coordination control strategies, can effectively reduce motor output torque fluctuations and improve corresponding speeds.

To reduce the fluctuation effect of the power output during switching, researchers have undertaken a lot of work on the boundary problem of mode switching. Wang [137] proposed an adaptive mode-switching strategy for multi-mode hybrid storage systems of electric vehicles, based on simulated annealing optimization. This strategy effectively minimized the occurrence of mode transitions while simultaneously preventing abrupt surges in battery power output. Zhou [58] designed a power-shift control strategy for a parallel hydraulic hybrid electric vehicle. The hydraulic pump/motor (HPM) was employed to provide torque compensation when the engine clutch was disengaged, maintaining power continuity and system stability. The linear quadratic regulator control strategy was used to control the process of engine clutch engagement to reduce the vibration of the transmission system. The influence of nonlinear factors (hydraulic hysteresis, oil temperature changes) has not been considered here. Wu [138] identified the driver’s required torque and divided the working mode regions. A coordinated control strategy for mode switching has been developed for three typical types. This paper reduced the frequency of mode switching and optimized system response through regional division. However, relying solely on torque demand partitioning patterns without considering differences in driving styles may lead to strategy failure. Zhang [139] introduced a novel control approach for mode transition that was specifically designed for dual-motor distributed coupling drive systems. Studies revealed that the maximum impact during the controller’s switching process measures 9 m/s³, falling below the suggested threshold of 10 m/s³. This control algorithm had high complexity and was sensitive to sensor synchronization and communication delay.

Model prediction methods have been extensively studied for energy management issues under established operating conditions. This approach offers the benefits of effective control outcomes and robust resilience, effectively addressing uncertainties, nonlinearities, and interdependencies during the transition process, but most actual driving conditions cannot be predicted in advance, and how to predict instantaneous energy distribution is an urgent problem to be solved.

5.3. Dynamic Collaborative Control Based on Torque Compensation

When the engine, motor, or hydraulic pump/motor is mode switching, the system inevitably experiences sudden changes in torque at the moment of switching, which requires the use of another power source to compensate for the sudden torque-change phenomenon. Wang [140] proposed a composite torque-coordinated control strategy to enhance the quality of mode transitions for power split hybrid electric vehicles and reduce torque fluctuations at the energy distribution interface of the coupling assembly by basic motor torque compensation. A composite coordinated control strategy was proposed, which includes a predictive controller, a motor torque-change-rate-limiting module, a fixed communication priority network transmission mechanism, and a fault-tolerant torque-coordination control module.

Zhang [141] implemented closed-loop regulation of driveline rotational dynamics through actuator velocity modulation strategies, establishing a hierarchical control architecture that enhances velocity synchronization precision. Hierarchical control architecture can simplify the design of complex systems and improve modularity. However, the interaction between the lower-level actuators and upper-level controls may introduce delays or errors, especially in dynamic conditions such as rapid acceleration/deceleration, and the real-time communication between layers needs to be strictly verified. Cheng [130] designed a disturbance rejection optimal control law of “disturbance estimation and disturbance optimal control” for the key problem of clutch sliding wear, as well as a shift strategy for parallel control of the clutch sliding wear stage and the driving torque recovery stage. The disturbance estimation in the literature depends on the accuracy of the model. In actual working conditions, underestimation may lead to a decrease in control effectiveness. Xue [142] primarily investigated optimization methodologies for drive system velocity regulation while developing synchronization protocols for motor–clutch interaction management, and proposed a proportional integral differential bangbang fuzzy composite intelligent algorithm for the motor speed control system. The proposed shift coordination control strategy outperformed the PID-based approach, cutting shift time by 35.7% and lowering clutch slip friction work by 19.2%. The composite calculation of PID and fuzzy control may increase algorithm delay, and its feasibility under a high-frequency shifting strategy needs to be further verified.

To suppress the impact vibration caused by mode switching in mode-switching strategy, researchers applied fuzzy theory to the switching process by using auxiliary components, and took into account the different response characteristics between the generator and the motor. Cui [131] proposed a mode-switching control strategy based on the difference in dynamic response characteristics between the engine and the motor, which included the engine torque change rate control, motor active speed control, and motor compensation clutch friction torque control. Results showed that the proposed mode-switching control strategy can effectively restrain the impact during the switching process. Fang [143] used secondary elements to compensate for the engine during vehicle driving to reduce the impact caused by torque fluctuations, so that the engine could work in the optimal working area. During the braking process, it effectively compensated for the loss of braking torque and suppressed the drastic changes in torque near the working point of the working condition. Tang [144] studied the impact problem during the switching process of parallel hybrid electric vehicles based on fuzzy theory. A fuzzy controller was designed for testing and analyzing the UDDS operating conditions, and the curves of engine torque change, motor torque change, and impact change during mode switching were studied (Figure 13).

In summary, most studies are based on predicting the real-time torque output of the engine and subtracting the predicted value from the required target torque, and the resulting difference is compensated by the motor for torque. If the clutch engagement or disengagement is involved in the mode-switching process, it is also necessary to control the torque transmitted by the clutch through clutch oil pressure and estimate the clutch torque. By coordinating the engine torque, motor torque, and clutch torque, the output shaft torque is not significantly changed, thereby solving the problem of unstable power transmission during mode switching.

The following deficiencies still exist in the current cooperative operation control strategies for mode switching, which require in-depth research:

(1)

The research on optimal energy allocation for unknown instantaneous operating conditions is not in-depth. For actual driving conditions [145] that cannot be predicted in advance, further research is needed on how to establish energy allocation based on unknown instantaneous conditions.

(2)

The research on dynamic collaborative control for mode switching is not mature. Although the motor torque response is fast to compensate for the torque of the engine or clutch, it is still necessary to comprehensively consider the clutch state and the coordinated operation strategy of the motor and engine to improve switching smoothness.

(3)

How to compensate for the torque of hydraulic pumps/motors with motors is a research gap. There are existing methods that use the motor torque’s fast responsiveness to compensate for the engine torque lag, but there is a lack of strategies for quickly compensating for the hydraulic pump/motor torque lag.

(4)

Switching between braking modes needs further research. Most of the literature has conducted much research into driver mode switching. However, hybrid vehicles have a large number of braking modes in actual operation, which has great development value for researching the recovery of braking energy.

6. Application of Electro-Hydraulic Hybrid System

Building upon EHHS, an innovative multi-energy integration system (MEHS) is recommended, which integrates a pair of electric motors with a piston pump/motor into a unified system. This advanced mechanism enables a seamless energy transformation between electrical, kinetic, and hydraulic power domains. This chapter begins by outlining the architecture and operational principles of mechanical-electro-hydraulic couplers (MEHC). Subsequently, the application of mechanical-electro-hydraulic power-coupled systems (MEHPCS) in vehicles is analyzed, and the effects of energy recovery and utilization are analyzed.

6.1. Mechanical-Electro-Hydraulic Coupler

To minimize the substantial torque spikes at startup and boost the operational responsiveness of EHHVs, a novel mechanical-electro-hydraulic coupler is designed from the perspective of vehicle driving. The innovative hybrid system integrates conventional permanent magnet rotational drive mechanisms with swashplate axial piston hydraulic units, enabling seamless bidirectional energy transformation among electrical, mechanical, and hydraulic domains. Its applicability spans across automobiles, construction equipment, agricultural machinery, machine tools, and more, with particular prominence as a power transmission component for electric commercial vehicles. Figure 14 illustrates the structure and working principle.

Yang [45,146] introduced an innovative vehicular architecture, integrating mechanical-electro-hydraulic coupling dynamics. Experimental validation demonstrated 14.418% and 21.174% reductions in MEH-PCEV battery depletion rates across distinct driving cycle configurations. A rule-based dynamic optimization EMS was established to achieve real-time control of the system’s energy allocation and the dynamic switching of working modes. Furthermore, this article designed a fuzzy controller to refine the control strategy to further reduce the peak torque of the motor. The results revealed that this scheme reduced the high-peak torque and improved the overall operational efficiency of the motor. Simultaneously, it led to an enhancement in the battery consumption rate.

To enhance the effectiveness of regenerative braking systems and avoid the impact on the hydraulic system, Meng [76] proposed a braking energy recovery strategy of the mechanical-electro-hydraulic coupling power vehicle that considered the optimal speed threshold. Results showed that when the vehicle speed exceeded 10 m/s and the energy recovery mode was switched, the overall recovery efficiency of the vehicle rose to 97.273%. Additionally, the SOC of the battery experienced a 0.14% increase.

The electromechanical hydraulic coupler integrates the motor and plunger pump/motor, greatly reducing the structural size, but the design difficulty also increases. Its input and output are coaxial, and there is only one output end. It only outputs power with a fixed transmission ratio, and cannot output variable proportions of mechanical and hydraulic power, according to the working conditions.

6.2. Mechanical-Electro-Hydraulic Power-Coupled System

Li [46,147] proposed a mechanical-electro-hydraulic power-coupled vehicle model, based on a planetary gear mechanism, to address the inability of a mechanical-electro-hydraulic coupler to output the required torque proportionally. This system is not only capable of driving vehicles forward but is also suitable for the working devices with hydraulic operation for construction machinery. The characteristics of this system are as follows:

(1)

A battery is incorporated alongside both high-pressure and low-pressure accumulators. These components are capable of adapting to diverse operational scenarios through mode switching. This leads to a substantial enhancement in the vehicle’s dynamic capabilities, facilitating the mutual conversion of mechanical, electrical, and hydraulic energies.

(2)

The powertrain integration relies on a planetary gear system to combine and distribute mechanical forces. By regulating the control motor and hydraulic pump/motor, the system achieves power division or coupling, addressing torque limitations, improving energy recovery, and maintaining motor stability.

(3)

It enhances power and improves energy recovery efficiency. Through the hydraulic power system, it can assist the power battery in increasing the instantaneous power of the system, briefly exceeding the original power limit of the drive system, and achieving electro-hydraulic synchronous assistance. By utilizing the battery and accumulators, the system can realize the recuperation and effective utilization of hydraulic potential energy and braking energy from the vehicle, improving the vehicle’s range and working time.

The system’s architectural composition and functional dynamics are depicted in Figure 15. The working modes are divided into the following six modes.

(1)

Stop mode: The vehicle remains stationary, with its velocity restricted under the parking configuration.

(2)

Electric driving mode (ED): The battery produces energy to rotate the motors, which transfers mechanical force through the planetary gear system, reduction gear, and differential, to propel the vehicle.

(3)

Hydraulic driving mode (HD): During energy conversion cycles, hydraulic oil is directed from the accumulator through the pump/motor functioning in power-generation mode. The resultant mechanical output interfaces with the drive axle via transmission components.

(4)

Electro-hydraulic driving mode (EHD): Collaborative operation between the motors and hydraulic pump/motor assembly generates torque. The planetary gear system enables real-time power coupling, while the driveline components transfer the combined power to the axle.

(5)

Electric regenerative braking mode (ERGB): The hydraulic pump transforms fluid energy into mechanical power, which is then channeled through the accumulator to produce the necessary braking torque.

(6)

Hydraulic regenerative braking mode (HRGB): Deceleration energy induces rotational actuation in the drive motor, facilitating mechanical-electrical energy transduction. The generated electrical output is redirected to electrochemical storage systems for subsequent utilization.

In this operational state, the pump/motor unit exclusively captures and reclaims energy generated during the deceleration process. When the hydraulic pump fails to deliver the necessary braking torque, due to insufficient power output, the control unit sets the valve opening range to −1, while the residual torque demand is compensated for by activating the friction-based braking mechanism.

For MEHPCS, when formulating energy management strategies, it is necessary to have a clear understanding of the conversion mechanism between mechanical, electrical, and hydraulic energy, as well as the relationship between the influencing factors of power synergy effects. Li [46] proposed a dynamic rule-based EMS that combined vehicle speed, accumulator pressure difference, braking strength, and demand torque, verifying the superiority of MEHPCS. Figure 15 demonstrated that the rule-based EMS [147] could achieve reasonable energy allocation and collaborative mode switching. To study the mechanism of the electro-hydraulic ratio and the overall performance, the electro-hydraulic ratio was further segmented. Electro-hydraulic energy partitioning analysis revealed that the optimal 0.5:0.5 ratio implementation achieved a 14.33% reduction in traction motor peak power and energy consumption, relative to conventional electric propulsion, coupled with 93.95% accumulator energy recuperation efficiency. The accumulator achieved an energy recovery rate of up to 93.95%. Finally, a multi-parameter EMS based on fuzzy rules was used to optimize and adjust the electro-hydraulic ratio in real time. The energy consumption of the battery based on fuzzy EMS was 2.09% lower than that of the battery with a fixed electro-hydraulic ratio EMS.

Furthermore, the system has the capability to link with hydraulic loads through hydraulic valve components, as shown in Figure 16 [148]. In contrast to the conventional electric excavator, the novel system has accomplished substantial energy conservation across four operating modes: ED, HD, HRG, and EGR. The potential energy recovery rate of this method was 92%. Most of the potential energy was stored in the battery through electrical energy. Moreover, the motor’s maximum torque in the new system was lowered by 66.7% compared to the conventional system. This reduction notably mitigated the occurrence of excessive peak torque during frequent starting and stopping, resulting in enhanced motor speed stability and the elimination of motor vibration concerns.

7. Prospects and Challenges

7.1. Prospects

The electro-hydraulic hybrid system is not only applied in vehicle power transmission systems but is also widely used in fields such as ocean wave energy utilization, vehicle suspension, and wind-power generation.

7.1.1. Wave Energy Converter

Among various renewable energy sources, wave energy holds considerable promise [149,150]. With its substantial power density and vast potential, it is estimated to possess a global capacity of 2 TW [151], thus being capable of playing a substantial role in addressing global energy requirements. The application of wave energy converters (WECs) for harnessing oceanic energy has been validated in both numerical simulations and physical experiments.

Drew [150] systematically classified contemporary wave energy converter (WEC) archetypes, delineating prevalent technological implementations. Secondly, some control strategies were considered to improve the efficiency of the point absorber type WEC. Henderson [152] proposed a new type of hydraulic power takeoff (HPTO) for the Pelamis WEC. The test results indicated that the combined efficiency of the main transmission exceeded 80% at full scale. Do [153] designed a suitable variable displacement hydraulic motor and generator. A new control strategy HPTO was proposed to improve regenerative power and energy efficiency. Dang [154] proposed an experimentally evaluated wave-energy converter, using HPTO for power conversion. A correction and control strategy for the pre-charging pressure of the accumulator was proposed to improve system performance (Figure 17).

7.1.2. Vehicle Suspension

The electro-hydraulic suspension in vehicles can not only reduce the vibration impact caused by road roughness but also convert this vibration energy into electricity and channel it into the battery via the motor.

Zou [155] developed a hydraulic interconnected suspension, utilizing regenerative damping technology. This suspension could not only attenuate vibration and maintain good riding comfort but also recover energy to energize the vehicle’s electrical components. Zhang [156] designed an electro-hydraulic semi-active shock absorber to collect the kinetic energy of the suspension. Wang [157] established a fully nonlinear physical “in service” model for the hydraulic shock absorber of the second suspension in rail vehicles. Chen [158] developed a dual-chamber hydraulic interconnected suspension system to boost the dynamic performance.

Cytrynski [159] introduced the active suspension system in a Mercedes Benz hydraulic pneumatic suspension. Equipped with continuously adjustable damping valves, the damper included hydraulic accumulators in its two working chambers. Abdelkareem [160] conducted a detailed study on a hydraulic damper suspension, based on energy collection. He pointed out that energy collection dampers based on hydraulic transmission were a promising track, especially in heavy-duty trucks and off-road vehicles. Sathishkumar [161] proposed a new method to harvest energy from dissipative active suspension through control valves. To achieve optimal comfort and grip, model predictive control was implemented for the hydraulic control valve (Figure 18).

7.1.3. Wind Turbine and Water Conservancy Generator

Wind power stands as a renewable and eco-friendly energy resource, and wind turbines serve as intermediary devices that transform wind energy into electrical power. In this research, hydraulic wind turbines combine the benefits of the inherent stability in hydraulic transmission systems, leading to a notable reduction in the effects stemming from wind speed fluctuations [152,162].

Laguna [163] proposed a hydraulic network device within a centralized power-generation wind farm. The primary function of the rotor nacelle module was to transport seawater to the hydraulic network, converting mechanical energy directly into hydraulic power and omitting any intermediary electrical transformation. Power generation was achieved at a central offshore installation via a Pelton turbine and transmitted onshore to traditional wind farms following a similar process.

Tidal turbines, as an effective technological means, have received great attention, as they contribute to achieving the goals of greenhouse gas emissions reduction and renewable energy production worldwide [164]. Fan [165] proposed a new offshore wind turbine that included fluid power transmission and energy storage systems. Seawater was sucked in through a variable displacement pump in the cabin. This variable displacement pump was directly connected to the rotor and used to drive an impact turbine installed on a floating platform. To ensure the smoothness and stability of the output power, a flexible energy storage system, capable of adjustable charging and discharging, was implemented.

In order to achieve smooth system switching and reduce vibration, researchers have proposed many control strategies. Li [166] proposed a smooth switching control strategy for the voltage source and current source dual-mode operation of doubly fed wind turbines, which solved the problem of large current surges and output power fluctuations during mode switching. Ai [162] studied the impact problem caused by underwater hydroelectric wind turbines under fluctuating conditions. A control method for pressure resonance was proposed, featuring real-time terminal impedance adjustment to suppress resonant frequency activation. Schulte [167] integrated a Takagi–Sugeno sliding mode observer into hydrostatic transmission wind turbines for actuator fault diagnosis and fault-tolerant control (Figure 19).

7.2. Challenges

Vehicles equipped with EHHSs, although possessing super strong driving and efficient recycling efficiency, still face numerous challenges that cannot be ignored. Because of the frequent switching of operating conditions caused by the complex and ever-changing actual road conditions, more complex EMSs need to be considered for multi-power-source EHHVs. Especially when switching from one operating condition to another, there is an urgent need for relevant research results on how to suppress the instantaneous torque fluctuations of components in the early stage. Moreover, there is a lack of quantitative research on the interaction between electro-hydraulic integration, vehicle economy, hydraulic performance, and electrical performance. Accordingly, the research in this article has certain guiding significance for the development and EMSs of EHHVs and other hybrid vehicles.

The uncertainty of driving conditions and the complexity of actual working conditions lead to a large number of anomalies in the data collected from real road conditions. To reduce the occurrence of abnormal phenomena, most studies have adopted the method of high-frequency noise filters to smooth the speed. This method still faces certain difficulties. In this research, the internet of vehicles information technology, based on multi-source working condition information, is the future development trend of automobiles. We can use the platforms for data mining to eliminate the randomness of a single vehicle operating for a period of working conditions.

The vehicle with the EHHS effectively combines hydraulic pumps/motors and motors, achieving the conversion of mechanical, electrical, and hydraulic energy. However, the high-efficiency area of electric power does not coincide with that of hydraulic power, leading to a system efficiency decrease. To maximize the efficiency zone of the entire vehicle, there is an urgent need for an energy management strategy to solve the problem of overlapping system efficiency zones.

Due to the fact that the EHHS involves two power sources—battery and accumulator—at the moment of mode switching, it will cause a certain impact on the motor, hydraulic pump/motor, and coupling mechanism. However, there are currently few studies on transient switching problems. The existing methods for studying vehicles’ transient switching problems are applicable to pure electric vehicles and plug-in hybrid vehicles, and do not involve hydraulic power systems. The above methods cannot be immediately applied to EHHVs. As a result, it is necessary to develop a transient switching control strategy that is suitable for EHHSs to reduce transient fluctuations in components and improve their service life.

Meanwhile, it is difficult to balance real-time control and optimization strategies. Under the premise of ensuring vehicle stability, combining offline and online driving mode recognition with vehicle EMSs can reduce mode-switching sensitivity and reduce vehicle energy consumption.

EHHS is mainly used in other fields, such as ocean energy conversion, wind energy conversion, and large-scale mining machinery. The current hydraulic and electrical technologies have become very mature, but there is still enormous potential for development in the realms of energy conversion and energy-saving technologies.

8. Conclusions

This research reviewed the development status, applications, and EMSs of EHHSs, and proposed the challenges they face. Based on the advantages and characteristics of the EHHS, the configuration characteristics, coupling mechanism, and coupling method were studied, and the existing application scenarios of the system on vehicles were introduced. Secondly, this research explored the EMSs and COCSs of existing hybrid vehicles. In addition, the focus was on a mechanical-electro-hydraulic power-coupled vehicle that combined electric motors, hydraulic pumps/motors, and a planetary gear mechanism to accomplish a mutual transformation of mechanical, electrical, and hydraulic energy. The principles and control strategy of the vehicle were described. Finally, it provided a brief introduction to the applications of EHHSs in other fields, such as wave converters, vehicle suspensions, wind power generation, etc. There are still enormous prospects and challenges in EMSs and optimization based on the EHHS. The specific conclusions are as follows.

The summary of EMSs for hybrid vehicles is shown in

Table 3. A summary of different EMSs.

No.	EMS	Complexity	Real Time	Efficiency	Hardware Requirement
1	Rule-based EMS	Low	High	Medium	Low
2	Fuzzy EMS	Moderately high	Medium	Medium	Medium
3	ECMS	Medium	High	High	Medium
4	MPC	High	Low	Global optimum	High
5	DP	Extremely high	Extremely low	Global optimum	Extremely high
6	Adaptive weight genetic algorithm	Moderately high	Medium	Multi-objective balance	Moderately high
7	Deep reinforcement learning	Extremely high	Medium	Great potential	Extremely high

. In terms of algorithm complexity, rule-based EMS has the lowest and highest real-time performance, but its efficiency is average. The real-time performance of DP is extremely low and only suitable for offline optimization. The efficiency is globally optimal but cannot be adjusted in real time, and the hardware requirements are extremely high, relying on offline computing resources. In terms of efficiency, ECMS has high efficiency, but it also increases hardware requirements. Although the DRL algorithm has extremely high complexity, it has great potential for improving efficiency. Therefore, in practical applications, it is commonly used in a combination, such as ECMS + fuzzy control to improve real-time performance, or DP offline training + MPC online optimization.

(1)

Research on the coupled configuration of the EHHS. According to the energy-transfer path and system connection method, dynamic coupling systems can be divided into series, parallel, and hybrid types. Different forms of hybrid power are clearly distinguished in terms of structure, and there are also differences in the electro-hydraulic coupling mechanism. The structure of the coupling system is different, and the form of power conversion and energy conversion efficiency will be affected. Therefore, studying the coupled configuration for EHHS has significant practical significance for different vehicle application scenarios.

(2)

Composite energy management strategies. Firstly, we developed a composite EMS to meet the demand for conversion among electrical, mechanical, and hydraulic energies. The existing research on EHHSs mainly focuses on the working state of motors, with less attention paid to hydraulic systems. Secondly, with the emergence of intelligent driving systems, the use of historical operating conditions data to achieve online predictions of driving modes can provide a reference for control in the future. Intelligent strategy development requires the use of big data and expert experience. We integrated methods such as driving pattern recognition, deep learning, and machine learning to learn and train targets, which required power, motor status, hydraulic system status, and vehicle status. We also established a multi-level response interaction model from target to behavior to analyze the relationship between the migration behavior of motors and hydraulic pumps/motors. The current trend of electro-hydraulic hybrid power systems is to combine driving condition prediction and Markov chain models to achieve dynamic energy consumption optimization. At the same time, artificial intelligence and adaptive algorithms are used to dynamically adjust energy allocation strategies to adapt to drivers’ habits and changes in road conditions, reducing gear shifting jerkiness. Finally, we integrated intelligent networking and vehicle road collaboration, combined with V2X data, predicting traffic information and optimizing energy flow in advance.

(3)

Collaborative operation control strategies. Multi-source hybrid vehicles require rapid switching between multiple power sources. It requires the system to have fast response characteristics while ensuring stability. So, the study of transient control strategies for mode switching is of great significance for electrohydraulic coupling systems. This control strategy can select objective functions and constraints to optimize the key parameters of the power system, using electric power to compensate for the slow response of hydraulic power and prevent the occurrence of hydraulic or electric power shaking when switching. Although the above control strategies have good effects, they impose a computational burden on the vehicle’s drive system and have poor real-time performance, making it difficult to apply in vehicles. Thus, the key parameters of the control strategy can be optimized and improved through the calculation results, based on optimized control algorithms. We determined the control rules for electro-hydraulic energy compensation, optimized the distribution mechanism and proportion between electrical and hydraulic energy, and achieved a globally optimal comprehensive collaborative operation control strategy. The multi-mode collaborative control strategy is developing towards intelligence, scenarization, and multi-objective collaborative optimization, and is using AI for algorithm-driven, real-time analysis of vehicle operating status, driving habits, and road conditions through machine learning, dynamically adjusting operating modes and power allocation. We quickly responded to mode switching and mitigated the mechanical impact on the components, and performed multi-source data fusion processing, predicted vehicle operating status and conditions in advance, and achieved full scene coverage and forward-looking control.

(4)

Wide application prospects. The characteristic of electric power is high energy storage density, while that of hydraulic power is high power density. From the perspective of application prospects, it is mainly applied in devices with high power density and high energy storage density. The system transforms marine kinetic power and atmospheric airflow into stored electricity, simultaneously mitigating suspension oscillations during energy conversion. Further research is needed on energy conversion and energy-saving technologies in the above application scenarios.

This study offered an essential overview of the management and advancement of EHHSs, and outlined EMSs and COCSs, providing a direction for future research on EHHSs. Research on electro-hydraulic hybrid systems will bring great advantages to hybrid vehicles and offshore applications. Firstly, the system has a high energy recovery and regeneration capacity, which can promote energy conservation and reduce emission pollution. Secondly, the high-power density can achieve fast charge and fast discharge and effectively reduce the production cost. Finally, research on the electro-hydraulic hybrid power system involves innovation in many technical fields, which could promote the development of related industries and provide a technical reference for other fields.

In addition, future research should prioritize the following key areas. Firstly, the configuration of the EHHS is not applicable to all scenarios. It is required to design a specific system structure in combination with vehicle characteristics and actual operating conditions. Secondly, the stable and collaborative operation of composite EMSs and mode switching are a major challenge for multi-source hybrid vehicles. This requires researchers to focus on the classification of vehicle driving modes and the detailed operational characteristics of the powertrain. Although some optimization EMSs can be utilized to optimize the power and economy for vehicles, it is still challenging for actual vehicle development. It is recommended to conduct in-depth research on the interrelationships between online and offline strategies and carry out targeted innovation in operational strategies.