New Trends and Challenges in Electric and Hybrid Electric Vehicles: Powertrain Configurations, Traction Motors and Drive Control Techniques

Abstract

The requirement of sustainable mobility and a clean environment has accelerated the development and adoption of electric vehicles (EVs) and hybrid electric vehicles (HEVs) as an alternative, practical and promising solution against conventional vehicles globally. Such alternative energy vehicles not only provide a critical solution to mitigate fossil fuel dependency and reduce greenhouse gas emissions, but also contribute to producing an energy-efficient transportation system. However, the operational performance, efficiency, and cost-effectiveness of EVs and HEVs are hugely dependent on their powertrain architectures, selection of traction motors and associated control techniques. This paper systematically compares major hybrid architectures: series, parallel, and series–parallel, plug-in, as well as battery and fuel cell electric vehicle platforms, highlighting trade-offs in component sizing, cost, and system integration complexity. The paper critically analyses traction motor technologies with respect to torque–speed characteristics, efficiency behavior, material constraints, and power density. A detailed comparative assessment of traction motor technologies is presented. Furthermore, classical and advanced motor control strategies, including field-oriented control (FOC), direct torque control (DTC), model predictive control (MPC) and AI-enhanced control frameworks, are evaluated with respect to transient performance, robustness, computational requirements, and scalability. The review identifies key technological milestones, emerging next-generation drive technologies, existing limitations, and unresolved research challenges. Finally, critical research gaps and future development pathways are articulated to support the advancement of high-efficiency, reliable, and cost-effective EV/HEV powertrain systems.

1. Introduction

The burning of fossil fuels in order to meet the energy requirements is a major source of greenhouse gas (GHG) emissions. GHG caused significant changes in the climate globally, which could be seen clearly in the form of glaciers melting and global warming [1,2]. As it is already estimated that global temperatures will rise by 2 degrees Celsius by 2050, the International Energy Agency (IEA) has developed major goals to reduce the average global temperature rise [3,4]. GHG emissions are expected to more than triple by 2050 and transportation accounts for about 25% of GHG emissions. Effective measures are essential to mitigate emissions from the transportation sector. Therefore, alternative fuels are being explored that can provide a clean energy source, minimize GHG emissions, and offer a cost-effective solution with high performance efficiency [5,6].

To address geopolitical concerns about fossil fuel availability, energy security, global climate change, and rising petroleum prices, transport electrification is regarded as the most appropriate solution. Furthermore, multiple sources of creating electrical energy, like renewable energy, provide for electrical energy diversification. The transportation industry is one of the largest sources of GHG and carbon dioxide emissions, as the energy conversion of internal combustion engines (ICEs) is between 12 and 30% [7]. EVs are the ideal solution as they have a high conversion rate and are more attractive for users over ICE [8].

Global EV demand is steadily expanding. Over 6 million battery electric vehicles (BEVs) and plug-in hybrid electric vehicles (PHEVs) were dispatched at a growth rate of 40% globally. In May 2023, China’s new EV passenger car sales grew by 60%, while EV sales in the United States increased by 48% and Europe-10 EV sales increased by 33%. Global EV sales prediction remains unchanged in April 2023, at 13.9 million units, with BEVs accounting for 72% of the total. The overall car industry recovery appears to be solid, with +23% year on year growth in the top-15 EV markets [9]. BYD is the leading brand in the global EV market in 2025. As seen in the year-by-year comparison, it took the most commanding sales position in the first half of 2025 with an increase of 33.3% [10]. HEV sales will reach thirty million globally by 2030, and 100 million by 2050 [11]. In Europe, 34% improvement in the EV market was observed in Q3 of 2025 [8]. Plug-in showed better sales progress compared to BEV in Q3 by a rise of 10% compared to Q2 of 2025. Over 25% of the new vehicles sold globally are EVs and the EV race is expanding to 39 countries where new EV manufacturers are contributing to 10% [12]. After a record-breaking Q3, a sharp decline in Q4 was observed in the US [13]. The year-to-year sales graph comparison of the USA is demonstrated in Figure 1.

Figure 2 illustrates EV sales growth comparison between 2024 and 2025 in Europe, the USA and China. This growth is continuously increasing, with full and mild HEVs growing by 14%, PHEVs by 5%, BEVs by 28% and overall EVs by 18%. However, shares of EV registrations in Europe, the USA and China for 2025 are demonstrated in Figure 3 [11,15].

Electrified transportation is a multidisciplinary field that includes the contributions and knowledge of different areas, as demonstrated in Figure 4. Traction motors play an important role in electric vehicle technology by improving performance and fuel efficiency [16,17]. The performance of the traction drive motor can either improve or degrade the performance of EVs. The environmental benefits of electric powertrains are seen and marketed as the true motivator for transportation electrification [18]. Manufacturers are constantly striving to create more efficient EVs [19]. EVs employed a battery bank to hold the charge and motors to power the vehicle, whereas ICE vehicles employed fuel and engines [20,21]. HEVs have high performance and efficiency, making them the most popular. Their large operating range and mature refueling infrastructure make them an alternative for the next generation [22]. There are various architectures of HEV propulsion systems that are already in place and providing future direction [23,24]. Most vehicles are charged while parked; the transition to EVs will place additional strain on the grid, particularly during peak hours [25]. To address this issue, an energy management system must be designed and executed. Various techniques have been presented; however, they are not physically implemented [26,27].

EVs employ a variety of motors; however, controlling an AC machine is a classic challenge [28]. Induction motors (IM) remain a smart alternative for EVs, even though the permanent magnet synchronous motor (PMSM) has been regarded as the dominant choice in recent years. Switch reluctance motors (SRMs) are also considered and employed in order to obtain a more affordable solution. Furthermore, it needed to be further improved for efficient performance. Although brushless direct current (BLDC) motors are highly efficient, they are preferred for lightweight vehicles. DC motors are the most basic motors, yet they are not appealing to manufacturers for EV applications these days. Double-stator and double-rotor motors are also under consideration for EV applications.

Motor speed control has a considerable impact on EV applications. Scalar control is a very basic approach, although DTC and FOC are superior speed controlling strategies. Intelligent controllers and MPC offer advanced speed control options. As technology advances in all dimensions, EVs are making strides towards smart cities through artificial intelligence and intelligent driving applications [29,30].

This paper is organized into six sections. Section 2 presents the classification of electric and hybrid vehicles, focusing on various powertrain architectures, their merits, limitations, and commercial utilization. Section 3 discusses the classification of traction motors, highlighting their applications and current challenges. Section 4 and Section 5 explore high-performance traction motor control techniques with an emphasis on artificial intelligence integration. Section 6 addresses the present challenges, proposes research recommendations for potential solutions, and provides the overall conclusion of the paper, respectively.

2. Classification of Vehicles

Vehicles are mainly categorized into internal combustion engine (ICE) based automobiles and EVs, either fully or partially electric. HEVs include full hybrids and plug-in hybrids, while all-electric vehicles (AEVs) consist of battery electric vehicles (BEVs) and fuel cell electric vehicles (FCEVs) as demonstrated in Figure 5 [31]. ICEVs suffer from inefficiency and harmful emissions, prompting a shift toward battery-powered EVs to enhance efficiency and reduce air pollution for a healthy and clean environment. EVs and HEVs offer several advantages such as higher efficiency, lower emissions, reduced noise, and conservation of petroleum resources for alternative applications [32,33].

2.1. Hybrid Electric Vehicle

In HEV design, the primary focus is to regulate power flow between the electric source and the load to reduce power losses. Both the traction motor and ICE are utilized for the vehicle propulsion, achieving performance comparable to ICE vehicles as shown in Figure 6 [34,35]. Enhance dynamic performance in HEVs, which relies on careful energy configuration and hybrid power in powertrain systems [36,37]. On the basis of electrification level, HEVs are classified in micro-hybrid EV (mHEV), mild-hybrid EV (MHEV), full hybrid EV (HEV), plug-in hybrid EV (PHEV), and extended range EV (EREV) [32,38]. Full hybrid configurations are further characterized as series, parallel, and series–parallel [39,40]. HEVs are preferred due to their fuel efficiency, ultra-low emissions and regenerative braking advantages [38]. However, a hybrid powertrain increases vehicle weight, manufacturing cost and maintenance complexity. To overcome these issues, hub motors are introduced, which are positioned directly at the wheel hub, eliminating conventional axles and gearsets. The electric axle (E-Axle) drive integrates the propulsion motor, inverter and transmission in a single unit, optimizing energy conversion, torque distribution and overall power efficiency [41]. The hybridization factor (HF) expresses the relationship between the maximum power of the electric motor (PEM) and the power of the ICE motor (PICE). HF for EV is 1, whereas HEV performance could be optimized at (HF = 0.3–0.5). After that, improving the propulsion system capability did not increase HEV performance [32].

HF = PEM/(PEM + PICE) = PEM/PHEV

(1)

2.1.1. Series HEV Systems

Series HEVs (SHEVs) are the simplest HEVs, consisting only of a battery and engine-generator set and propulsion via a traction motor, as demonstrated in Figure 7 [42]. ICE charges the battery bank while consuming minimal fuel. The drive train is employed in hybrid, engine, electric, charging, or regenerative braking mode. EREV is a prime example of SHEVs. SHEVs are appropriate for short-distance driving and stop-and-go traffic [43]. The advanced architecture provides optimal power control between motor and generator with a simple drive train structure without gears and clutches [44]. However, dual power conversion, huge traction motor size and large battery banks are the concerns. SHEVs operational characteristics, present challenges and system applications are demonstrated in

Table 1. Characteristics of series HEV configuration [9,14,27,32].

Advantages	Disadvantages	System Application
The traction drive line is optimized and efficient. Small and efficient engine. Zero-emission operation is possible. The installation flexibility of machines. No multi-step transmission.	Energy loss due to multiple energy conversions. Large traction motor drive. Increased cost because of the additional generator.	BMW i3 Fisker Kerma Tesla Buses Mercedes Citaro bus Tesla ultra-light rail

.

2.1.2. Parallel HEV Systems

Parallel HEV configurations are propulsion systems that connect both the ICE and the battery in parallel with the traction motor, as demonstrated in Figure 8. However, a switch allows both systems to operate concurrently. The battery runs the vehicle at moderate speeds, while the ICE runs the vehicle at high speeds. Therefore, it provides a compact system with superior dynamic performance. The recent advancement in compact integration of motor transmission coupling structure provides torque blending, regenerative braking while reducing drivetrain complexity [44]. However, difficult control, complex space packaging, and high system cost are the concerns. Parallel HEVs’ operational characteristics, present challenges and system applications are demonstrated in

Table 2. Characteristics of parallel HEV configuration [9,13,27].

Advantages	Disadvantages	System Application
Increased efficiency as a result of less power conversion. No additional generator. High speed due to two power sources. Zero emission is possible. One machine is required for hybrid operation. Less space is required for additional installation.	Complex space packaging. Expensive system. Complex control.	Ford Escape Hybrid SUV Honda insight Mercedes-Benz 400 Honda Civic Hybrid. Honda Accord. Lexus Hybrid SUV.

.

2.1.3. Series–Parallel (SP-HEV)

Series–parallel HEVs provide the capabilities of both series and parallel HEVs with the help of a planetary gearbox, as shown in Figure 9 [45]. At low speeds, electric motors generate the most torque, whereas at high speeds, ICE takes the lead. This setup improves the vehicle’s power capability, low emissions, and fuel efficiency. An advanced hybrid drivetrain provides fuel efficiency and flexible power flow [44]. However, control complexity, high system cost, and complex drive train layout are the concerns [46,47]. SP-HEVs’ operational characteristics, present challenges and system applications are summarized in

Table 3. Characteristics of SP-HEV configuration [39,49,50].

Advantages	Disadvantages	System Application
Vehicle power capability is improved. Torque is transmitted directly to the drive wheel. Better fuel efficiency. Zero emission is possible. Multi-step conversion is not required. It can be switched b/w series and parallel.	Two motors are required Cost increases System complexity	Toyota Prius. Toyota Camry Hybrid Hyundai Ioniq

. In addition, several power-split modes are addressed in [48].

Comparison of different operational modes, which are summarized in

Table 4. Comparison of operational modes in series–parallel HEV [40,42,47,49,50].

Modes	ICE Dominating	EM Dominating
Starting phase	Electric motor and battery, then ICE	Electric motor and battery, then ICE
Full throttle acceleration	Both active, but ICE provides dominant power	Both active, but EM provides dominant power
Normal drive	ICE	EM
Braking/Deceleration	EM acts as a generator to recharge batteries	EM acts as a generator to recharge batteries
Charging during driving	ICE + generator charge batteries	ICE + generator charge batteries
Vehicle at rest	ICE drives the generator to charge batteries	ICE drives the generator to charge batteries

.

2.1.4. Plug-in HEV

Plug-in HEVs employ an external source to charge the battery. When the battery is depleted, the engine will resume functioning [51]. If the battery is charged, it has good fuel efficiency and low emissions, but as the battery discharges, the fuel consumption and CO₂ emissions increase. PHEVs are gaining popularity among consumers as their sales growth is rising by 40% in 2023, by 16% in 2024 and 20% in 2025 [10,52]. PHEVs are the electric drive train as the primary source of propulsion and energy. It is critical to pick the appropriate power train to drive plug-in vehicles to improve efficiency and mileage while reducing emissions [53]. Plug-in drawbacks are relatively large in size, which means the fuel efficiency is not good for highway travel or when the battery is not fully charged. However, Lithium-ion batteries are now extensively employed in Plug-in EVs and other types of EVs to boost performance [4,54]. Plug-in-HEVs operational characteristics, present challenges and system applications are summarized in

Table 5. Characteristics of the plug-in HEV system [9,47,55].

Advantages	Disadvantage	System Application
Low emission Low fuel consumption Extended electric drive range	Dependent on the grid High initial cost Complex structure	Hyundai Ioniq Plug-in Volvo XC90 T8 Mini Countryman PHEV Porsche Cayenne e-Hybrid Volkswagen Phideon

.

2.2. Pure Electric Vehicle

The vehicle propulsion mechanism in pure electric vehicles is powered by electrical energy. It may also use renewable charging sources such as solar cells, wind energy, or other alternatives. This helps reduce dependence on the grid power supply [56,57,58]. Manufacturers are employing regenerative braking technologies during braking and as well as when traveling freely under the force of gravity [59]. The benefits include zero emissions, high mileage, very low noise pollution, compact size, and excellent efficiency. Development of unmanned, self-driving, and intelligent electric vehicles with one or more power train driving motors contributes more in the future [60,61]. However, the concerns are low load-bearing capacity, slow speed when compared to HEV and ICE vehicles, complex structure, and high cost owing to a huge battery bank.

BYD’s sales more than doubled to 641,000 units, propelling it to the top of the global sales rankings. Among BEVs, Tesla continues to lead by a significant margin in 2024. However, it is the only brand whose sales dropped to 13.1% in 2025. FCEV and ICEV sales are down by −9% and −16% in 2025, respectively, from 2024. The comparison of each arrangement is presented in

Table 6. Comparison of the global passenger EV market shares [62,63].

Global EV Market Share Up to Q3 2025
Brands	Q1 2024	Q2 2024	Q3 2024	Q4 2024	Q1 2025	Q2 2025	Q3 2025
BYD Auto	15%	17%	16%	16%	15%	18%	16%
Tesla	20%	17%	17%	14%	12%	12%	13%
Geely Holdings	8%	8%	9%	9%	11%	11%	10%
Others	57%	58%	59%	61%	61%	60%	61%

. Plug-ins (BEV and PHEV) accounted for an increase of 34.5% in 2025 H1. BEVs made up 10.4% of the market, while PHEVs made up 4.4%. Norway had the greatest PEV share in new light vehicle sales year to date (80.6%), followed by Hong Kong (68.5%), Iceland (54.9%), Sweden (52.3%), and Finland (46.8%) [9,62].

In 2024, BEV continues its selling trend by 12% YoY in Q4 and BYD’s sales overtake Tesla by 13% YoY. In China, the growth rate is reaching 56% in Q2, while America and Germany are following the increasing trend. The best-selling car in the world ranking for the year 2024 is BYD Song, led by 467,750 units but BYD Seagull leads in 2025, Tesla Model Y securing +3.6% market leadership in 2024, despite a drop in 2025, while Geely Holdings is in third place [9,10]. BYD and Tesla were still dominant at the end of 2025 [12].

2.2.1. Battery Electric Vehicle

A battery electric vehicle has three major components: an electric motor, a controller, and a rechargeable battery bank or super capacitor (SC), as demonstrated in Figure 10 [64]. The propulsion system is powered entirely by electrical energy. The electric motor is the primary source of EV movement and torque [65]. The recent development provides a compact module by integrating motor, inverter and transmission in an electric drive unit to enhance the efficiency. The adoption of a multi-level inverter mitigates switching losses and improves further performance [66]. The transition of operating voltage from 400 to 800 reduces current and minimizes power losses while enabling faster response and enhanced power density [67]. Due to the enormous battery bank, size and cost increases. It requires external charging, whereas HEVs have internal charging capabilities [68]. Furthermore, the charging infrastructure is not yet mature enough all over the world, despite rapid advancement in developing countries.

Level 1 charging provides slow charging; however level 2 and level 3 charging stations are also advancing with fast charging topologies [69,70] and a fully charged battery provides approximately 200 km [71]. However, BEVs operational characteristics, present challenges and system applications are summarized in

Table 7. Characteristics of BEV [9,13,62,72].

Advantages	Disadvantages	System Application
Zero emission. Good fuel efficiency. Reasonably high-power density.	Bigger in size and high cost. Battery dependency. Immature charging infrastructure.	Audi e-tron BMW i3, i7 Cadillac Lyriq Chevrolet Bolt EV Ford Mustang Mach-e

.

2.2.2. Fuel Cell Electric Vehicle

Fuel cell electric vehicles run on hydrogen as a fuel. Compressed hydrogen is stored in a fuel stack, which employs oxygen from the air to generate electrical energy, water and heat as a byproduct. Figure 11 shows the basic block diagram of FCEV. Platinum is employed as a catalyst to shorten the reaction time [73]. FCEV has high power density and fast refueling, which makes it suitable for long-range applications [74,75]. However, the conversion rate of hydrogen is about 40–60% as well as reliance on hydrogen infrastructure limits practical application [76]. Similarly, the complexity of the system, greater starting costs, fuel costs, reliability, and cyclic life are the challenges. FCEV has substantially lower specific power and lower starting performance; therefore, it is not sufficient by itself [77]. It requires batteries or a supercapacitor in hybridization, which increases cost and size. Therefore, A hybrid optimization technique is proposed for dynamic load sharing to tackle this issue for FCEV [78,79]. However, emerging electrified platforms accommodate multiple energy sources within a dedicated propulsion architecture to improve system efficiency [80]. Another method to deal with non-linearity and power losses based on fractional sliding mode control (SMC) design is proposed for FCEV [77].

As pure hydrogen is not available for refueling, the following are the commonly used compounds in FCEV. Direct methanol fuel cell (DMFC), alkaline fuel cell (AFC), phosphoric acid fuel cell (PAFC), solid oxide fuel cell (SOFC), and other types of compounds are commonly utilized in FCEV. A comparison of FC compound parameters is demonstrated in

Table 8. Comparison of FC compound parameters [81].

Compound	Fuel Type	Power Density (W/m3)	Temperature (°C)	Efficiency (%)
DMFC	Methanol	1500–3500	<100	30–40
PAFC	Hydrogen	800–2500	150–220	40–55
AFC	Hydrogen	1000–3000	50–150	40–60
PEMFC	Hydrogen	3500–6500	50–100	45–60

.

PEMFC is widely utilized in FCEV. It has a small size, a good start at low temperatures, a simple structure and maintenance-free operation, a reasonably high-power density, and the capacity to run under harsh conditions [82]. However, FCEVs operational characteristics, present challenges and system applications are summarized in

Table 9. Characteristics of FCEV [83].

Advantages	Disadvantages	System Application
Low emission. Good fuel efficiency. Hydrogen as fuel	Additional chemicals are required. Low starting performance. Lack of infrastructure to produce pure hydrogen.	Toyota Mirai BMW i8 Audi H-tron quattro Mercedes-Benz GLC F-cell

.

A detailed comparison of ICE, BEV, HEV, PHEV and FCEV is demonstrated in

Table 10. Comparison of ICEV, BEV, HEV, PHEV and FCEV [14,15,83,84,85].

Characteristics	ICEV	BEV	HEV	PHEV	FCEV
Propulsion System	IC engine	Electric Motor	Electric Motor and IC engine	Electric Motor and IC engine	Electric Motor
Energy Storage	Fuel tank Gas cylinder	Battery Super Capacitor Flywheel	Fuel tank, Battery Super Capacitor Flywheel	Fuel tank, Battery Super Capacitor Flywheel	Fuel cell Super Capacitor Battery
Energy Source Infrastructure	Fuel Station	Electric grid charging station	Electrical charging station Fuel pump	Electrical charging station (optional) Fuel pump	Hydrogen PEMFC, AFC, DMFC, SOFC
Advantages	Fully matured High performance Reliable Availability of refueling infrastructure	Zero emission. High efficiency Low noise Independent of gasoline Commercialized	High performance Reliable Low emission Two sources of energy Higher fuel economy Durability Commercialized	Low emission Low fuel consumption Extended Electric drive range Smooth operation Capability of V2G or G2V	Ultralow emission. High energy efficiency Independent of gasoline Reliable
Disadvantages	Harmful emission Poor fuel economy Less efficient Comparatively bulk	High cost Poor dynamic response Long charging time Immature charging infrastructure Limited driving range	More complex due to two energy sources Bulky High cost Size and weight increased	Dependent on the grid High initial cost Complex structure Charging station infrastructure Battery technology Size and weight increased due to batteries and ICE	Pure hydrogen availability Slow dynamic response High cost Immature charging infrastructure Sophisticated electronic controller Conditioning, refilling and storage of hydrogen
Comparison 2021 vs. 2022 2023 vs. 2022 2024 vs. 2023 2025 vs. 2024	Still most dominant Sales declined by 16% ICE also increased +15% Decreased by 9% Decline by 18.7%	Attracting users BEVs grew by 75% BEV increased by 49% BEV increased by 9% globally Increased by 17%	Trending Full Hybrids grew by 14% HEVs increased 33% Increased by up to 30% Increased by 9.8%	Most trending EV PHEVs grew by 37%. PHEVs increased by 55% About 25% increases Increased by 20%	Decreasing trend FCEV declined by 9% FCEV decreased by 30% in Q1 2023 FCEV degrowth by 36.4% Declined by 27.2%

.

2.3. Batteries in Electric Vehicles

The battery system serves as the core energy source in pure EVs, fundamentally determining driving range, power capability, efficiency, and operational safety. Unlike ICE-based systems, EV propulsion relies entirely on electrochemical energy storage, making the battery pack a critical subsystem rather than a passive component. It typically consists of interconnected cells arranged in modules, integrated with a battery management system (BMS), thermal management unit, and protection circuitry. The coordinated operation of these elements ensures reliable energy delivery under diverse driving conditions.

2.3.1. Role of Batteries in EV Powertrain

Within the EV powertrain, the battery pack supplies DC power to the inverter, which drives the traction motor. It must support both continuous energy demand during cruising and transient high-power requirements during acceleration. Additionally, the battery enables regenerative braking by absorbing recovered energy, thereby improving overall system efficiency [86].

The BMS performs essential monitoring and control functions, including estimation of state-of-charge (SOC) and state-of-health (SOH), voltage balancing across cells, and temperature supervision. Accurate estimation of these parameters is necessary to prevent overcharging, deep discharging, and thermal instability, which can significantly degrade battery performance and lifespan [87].

From a structural perspective, battery placement usually occurs under the vehicle floor to enhance stability by lowering the center of gravity. However, this configuration imposes constraints on packaging, thermal dissipation, and crash safety design.

2.3.2. Battery Chemistries for EV Applications

Lithium-ion (Li-ion) batteries currently dominate EV applications due to their high energy density, long cycle life, and relatively low self-discharge rate. Among various Li-ion chemistries, the most widely adopted include:

• Lithium Nickel Manganese Cobalt Oxide (NMC): Offers a balanced combination of energy density, power capability, and durability, making it suitable for a wide range of EVs.
• Lithium Iron Phosphate (LFP): Characterized by superior thermal stability, enhanced safety, and longer cycle life, though with comparatively lower energy density.
• Lithium Nickel Cobalt Aluminum Oxide (NCA): Provides high specific energy and is commonly used in long-range EVs, albeit with stricter thermal management requirements.

Recent advancements in Li-ion technology have focused on improving electrode materials, electrolyte formulations, and cell design to enhance performance metrics and safety. Beyond conventional Li-ion systems, next-generation battery technologies such as solid-state batteries, lithium-sulfur, and sodium-ion batteries are being actively investigated. Solid-state batteries, in particular, offer the potential for higher energy density and improved safety by eliminating flammable liquid electrolytes [87].

2.3.3. Key Performance Metrics for EV Batteries

The suitability of a battery for EV applications is determined by several critical performance parameters:

• Energy Density (Wh/kg): Directly influences driving range and vehicle weight.
• Power Density (W/kg): Determines acceleration capability and regenerative braking performance.
• Life Cycle: Reflects durability and long-term cost-effectiveness.
• Charging Rate Capability: Defines the feasibility of fast charging without excessive degradation.
• Thermal Stability: Ensures safe operation across varying environmental conditions.
• Cost ($/kWh): A key factor affecting commercial viability and large-scale adoption.

These parameters are inherently interdependent, requiring careful trade-off analysis during battery selection and system design as shown in

Table 11. Comparison of batteries [88,89,90].

Parameters	Li-Ion	LFP	NMC	Solid State Battery
Energy Density (Wh/kg)	150–250	120–200	200–300	400–500 (target)
Power Density (W/kg)	1800	1000–1500	1500–4000	2000–5000
Charging Speed	Moderate–fast	Moderate	Fast	Very fast (future potential)
Life Cycle	500–1000	1500–3000	500–1250	1000– 3000 (expected)
Thermal Stability	Moderate	Very high	Moderate–low	Very high
Key Materials	Graphite + Li-metal oxides	Iron, phosphate	Nickel, cobalt, manganese	Lithium metal + solid electrolyte
Cost ($/kWh)	Medium	Low	High	Very high (currently)
EV Application	General EVs	Buses, entry EVs	Long-range EVs	Future EVs

. For example, high-energy-density chemistries often exhibit reduced thermal stability, necessitating advanced cooling strategies and control mechanisms.

2.3.4. Challenges in EV Battery Technologies

Despite significant technological progress, several challenges continue to hinder the widespread adoption and optimization of EV batteries:

• Performance Degradation: Battery aging mechanisms, including solid electrolyte interphase (SEI) growth, lithium plating, and electrode degradation, lead to capacity fade and increased internal resistance over time.
• Thermal Management: Maintaining uniform temperature distribution is critical for performance and safety. Non-uniform thermal profiles can accelerate degradation and increase the likelihood of thermal runaway.
• Safety Risks: Lithium-ion batteries are susceptible to thermal runaway under mechanical, electrical, or thermal abuse conditions, necessitating robust monitoring and fault mitigation strategies.
• Fast Charging Limitations: High charging rates can induce excessive heat generation and accelerate degradation, requiring optimized charging protocols.
• Material Constraints: The reliance on critical raw materials such as lithium, cobalt, and nickel introduces concerns related to cost, availability, and environmental impact.
• End-of-Life Management: Efficient recycling and second-life utilization remain essential for improving sustainability and reducing environmental burden [88].

2.3.5. Emerging Trends and Research Directions

Ongoing research efforts are focused on addressing the aforementioned challenges through innovations in materials, system design, and control strategies. Artificial intelligence-based BMS approaches have demonstrated improved accuracy in SOC and SOH estimation, enabling predictive maintenance and adaptive control [91].

Furthermore, advancements in battery thermal management systems, including liquid cooling and phase-change materials, are enhancing safety and performance under high-load conditions. The development of solid-state batteries represents a promising direction for achieving higher energy density and improved intrinsic safety, although challenges related to scalability and manufacturing persist.

Another emerging approach involves hybrid energy storage systems that combine batteries with supercapacitors to handle peak power demands while reducing stress on the battery. Additionally, second-life applications of EV batteries in stationary energy storage systems are gaining attention as a means to extend battery utility and improve lifecycle sustainability.

3. Classification of Traction Motors

The role of the traction motor is crucial for the advancement of the electrified transportation system. Electric motors are widely employed not only in domestic and industrial applications, but also play a vital role in the development of modern EVs and HEVs. Figure 12 and Figure 13 present various types of AC and DC traction motors [92]. However, Figure 14 presents cross-sectional views of key traction motor technologies: IM, PMSM, and SRM. Advances in motor design, materials, electronics, and control techniques continue to enhance efficiency and reliability across various applications [72,93].

The choice of an electric propulsion system depends on factors such as vehicle constraints, driver demands, and energy sources [31,80]. Various traction motors are available today; however, the selection of any particular motor for an electric propulsion system depends on factors such as simplicity, durability, lightweight design, compact size, cost-effectiveness, flexible control, fault tolerance, low noise, high efficiency, and reliable performance across diverse driving conditions [79,80]. Additionally, market acceptance, influenced by the availability and cost of compatible power converters, plays a very important role in the motor selection.

3.1. DC Motor

The technical and economic advantages of DC motors historically made them the preferred choice for most EV and HEV manufacturers initially [94]. Their widespread adoption was driven by simple speed control, technological maturity, lower manufacturing costs, and superior speed regulation compared to other electromechanical conversion devices [95]. However, the use of commutators and brushes in DC motors reduces reliability, rendering them unsuitable for maintenance-free and high-speed applications like EVs and HEVs. Additionally, issues such as electromagnetic interference, bulky construction, low efficiency, and low specific power density further limit their suitability for EV and HEV applications [96]. Compact design and high efficiency are crucial for the development of an electrified transportation system; therefore, BLDC motors are under consideration.

3.2. Induction Motor

IM is one of the most suited and mature electric propulsion systems due to its high dynamic performance. IMs can extend their constant power range up to 4–5 times the base speed, meeting traction requirements. However, high-speed operation is constrained by pullout torque limitations [95]. Figure 15 shows a typical speed–torque characteristic of an IM and its variations with rotor speed.

IM propulsion systems have various advantages, including minimal noise and ripples, mature and efficient power electronic drivers, reliability, and simplicity of design. Variable frequency and voltage control can produce high starting torque and low starting current, while high efficiency in IMs can be achieved through minimum slip control [96,97]. Pulse width modulation (PWM) is typically employed to control the speed of the motors. However, it generates unwanted common-mode voltage, reducing motor life [98]. IM has achieved efficiency of 90%; however, thermal losses, larger size, and heavy weight are concerns that reduce the manufacturer’s attraction [41]. PMSMs challenge IM dominance with 95% efficiency, higher power density, and lower rotor losses.

To enhance IM drive efficiency, several control techniques tailored for HEV applications have been proposed [78]. An optimization technique is proposed for deep regulation of magnetic flux to improve voltage drop and efficiency [79].

The development of better control techniques eroded IM’s cost advantage. EV vane makers and Tesla had already deployed PMSM in their new models. IM’s speed control is broader and much easier to achieve than PMs. Due to the absence of PM’s high energy density, IM has a low torque performance. IMs operate with less than 3% slip, minimizing rotor core losses and high efficiency [88]. However, maximum efficiency is achieved when iron and copper losses are balanced [89]. Nevertheless, core and resistance losses are the significant contributors to IM efficiency when compared to stray losses [98,99]. The following

Table 12. Characteristics of IM [99,100,101].

Types of Losses	Proposed Method	Adopted Technology		Advantage
Copper loss	Increase conductivity	Replacement of an aluminum conductor with a copper conductor		Copper loss reduced Efficiency improved
Iron losses	Improve model calculation	3D finite element analysis		Calculation efficiency improved
		A combination of vector control and the Fe loss model		Independent control of torque and speed Fe loss reduced
		Develop a controller on the basis of a stationary coordinate system		Fe loss reduced Less calculation Simplified model
	Fe loss model-based control technology	Combination of DTC and Fe loss model		Fe loss reduced Torque response is faster
		Combination of sliding mode control and Fe loss model		Robustness improved
	Search method	Iterative flux for input power reduction		Fe loss reduced Motor parameters are not sensitive
	Combine search and model	Initially, use the Fe loss model for the calculation of the magnetic flux approximate value, then go for the optimal value		Fe loss reduced Comparatively high accuracy Calculation speed increased
Structure losses	Employed Method	Efficiency (%)	Torque Density (Nm/L)	Advantage
Prove the structure design	Increase axial length	88	Not given	Efficiency improved Implementation is easy
	Skew rotor structure	87.3 86.4	33.17 36.17	Efficiency improved Torque performance improved
	Multi-objective optimization	89 86.5–87.7	22.4–39.1	To maximize efficiency, other performances are taken into account

provides specifics about each loss.

3.3. PMSM

Since the last decade, PMSM has been the most demanding motor in the EV sector. Permanent magnets could be surface-mounted on the rotor or interior-mounted, with the magnet inside the rotor [102,103]. Advantages of PMSM include excellent efficiency, high starting torque, compact size, smooth control, decreased noise, low power consumption, good dynamic performance, and dependability [104,105]. Furthermore, low rotor losses and high-power density make them the most suitable motor [106,107,108]. The smaller size enables greater dependability, effective heat dispersion into the atmosphere, and maintenance-free operation. However, the motor cost is high due to rare-earth permanent magnets and necessitates the use of a motor control drive [109]. Sensored and sensorless control strategies are employed to achieve further improvement. A hybrid smooth switching, speed controlling strategy from low to high speed is discussed [110,111]. The impact of the air gap of PMSM is studied for shaft torque [110]. A fault-tolerant issue is studied by increasing the power electronic converter rating due to high-speed control and management of back emf.

As rare-earth magnets are costly due to their non-renewable nature and low productivity, alternative magnets or rare-earth-free designs are under consideration. However, the lower power density of rare-earth-free magnets remains a challenge. Ongoing studies aim to improve the power density and efficiency to make them more viable for advanced applications. Ferrite PM, which has a very low material cost, is one of the least expensive alternatives. A roadmap has been proposed to improve the efficiency and power density with reduced cost [112]. The following is a comparison of two rare-earth magnets (Samarium cobalt and Neodymium) and two rare-earth-free magnets (Alnico and Ferrites), as demonstrated in

Table 13. Comparison of magnetic materials [114,115].

Property (Symbol, Unit)	Alnico	Ferrites (Ceramic)	Samarium Cobalt	Neodymium
Coercive Force (Hc, kA/m)	37–143	0.23–0.41	480–840	760–1030
Density (d, g/cm3)	6.8–7.3	4.9	8.4	7.4
Electric resistivity (r, Ω cm)	(50–75) × 10−6	10−6	(53–86) × 10−6	160 × 10−6
Max. Service Temperature (Tmax, °C)	450–550	800	300–350	150
Max. energy Product ((BH)max, kJ/m3)	10.7–71.6	8.35–31.8	130–240	220–336
Price, USD/k	58	7.1	100	75
Remanence (BR, T)	0.7–1.28	0.23–0.1	0.83–1.16	1.00–1.41

[113].

3.4. BLDC Motors

BLDC motors are electronically commutated; they do not require brushes, slip rings, or a physical commutator [116]. The windings are typically concentrated in nature and permanent magnets are placed on the rotor. A multi-step switching sequence is employed to generate a trapezoidal waveform. The back emf is employed to adjust the speed of the motor. Hall sensors are employed to regulate the speed and position of the rotor. BLDC motors have a very high efficiency, making them the best choice for electric bikes and rickshaws [117]. Furthermore, advantages include compact size, high efficiency, high starting torque, low power consumption, and reduced thermal losses [118]. However, the disadvantages of BLDC motors are the cost of neodymium magnets, magnetic field weakening in the in-runner motors and soft starting due to the high starting torque. An online monitoring algorithm is proposed to improve the reliability of the BLDC motor [119].

3.5. Switched Reluctance Motor (SRM)

SRM is described as an electronically commutated variable reluctance motor that employs reluctance torque to function. The stator has a single winding that is concentrated in nature, with no PMs in the rotor. Typically, stator poles have parallel sides, which means the edges are straight and equidistant. SRM operates on the variable reluctance concept, which means that the rotor will always follow the path with the least reluctance. The excitation sequence of the stator poles ensures that the SRM moves continuously. The number of poles on the rotor is mostly responsible for SRM’s outstanding performance. Furthermore, steady torque and power are impaired at high speeds. As a result of its simplified structure and low cost, the SRM is a highly suitable motor for electric vehicles. Additional benefits include high starting torque, good efficiency, a wide speed range, and a high torque to inertia ratio [120]. However, complex excitation circuitry, high torque ripples, iron core losses, noise, and increased production costs are the concerns [72,121,122]. Newly proposed C-core hybrid SRM topologies and optimized magnet placement strategies enhance torque density while keeping structural robustness and low cost. Further studies are being conducted on the reduction of high torque ripples to strongly utilize these motors in electrified vehicles [123,124]. Comparison of torque ripple reduction strategies in SRM is demonstrated in

Table 14. Comparison of torque ripple reduction strategies in SRM [122].

Method	Employed Techniques	Advantages	Disadvantage
Increase the number of rotor poles	Increased number of poles of the rotor	Average torque increased Torque ripple reduced Copper loss reduced	Complicated rotor configuration. Rotor material increased
Pole shape	Modification in rotor shapes, pole arc, pole shoe and air gap	Torque ripple reduced Efficiency increased High-speed performance	Complicated optimization for offline calculation
Modulation in current and angle	Turn-on and -off angles are optimized	Efficiency improved Torque–speed capability enhanced Torque ripple reduced	Limited current and torque control Large memory required to store current profiles
ATC and DTC	The hysteresis controller regulates the torque with online estimation	Torque ripple reduced Direct control of the instantaneous torque	Machine parameters prior knowledge required
Torque sharing function (TSF) method	TSF profile definition Implement hysteresis control with current reference from torque reference	Controlled torque easily Torque waveforms determined Smooth torque over a broad speed range	i-T-θ characteristics needed offline designed torque waveform
Feedback Loop (FBL) control	Transformation of a non-linear model into a linear model	Torque ripple reduced FBL is free from the non-linear term Provide the required decoupling of the current	Complex linearization algorithm. No-adaptability of uncertain parameters changes
Iterative learning control (ILC)	Current compensation to the phase current reference is added for current tracking	System parameter identification not needed Perfect current tracking on different conditions is achieved	Degraded performance to transient Complex learning control law Iteration cycle restricted
Intelligent Control	Offline and Online optimization of phase current	Strong self- learning Torque ripple reduced Adaptive ability Independent of machine parameters	Complex computational algorithm

.

3.6. Advanced Motor for EVs and HEVs

To boost torque density, a new breed of electric machine, known as a double-stator (DS) machine, is proposed [122]. The innovation of the double-rotor electric machine is developed from research on HEV power split devices [122]. Existing power split systems use mechanical gearboxes as power transmission devices, which are inefficient, noisy, and vibrational. Therefore, double-rotor electric machines can efficiently handle the problem since they employ magnetic force to convey torque and power, eliminating the need for physical touch. However, there is still a long way to go before the deployment of double-rotor and DS electric machines commercially. Their mechanical construction is complex and their power factor is low when compared to single-rotor electric machines. In some ways, these new topologies have a compact and complicated structure that is suited for EV powertrains as shown in Figure 16 [125]. DS IM provides more robustness and reliability with SMC [126]. Design analysis of an oil pump motor with PM is discussed to achieve maximum efficiency by using a different rotor [127]. Multiple criteria of different motors are discussed and analyzed [128].

Table 15. Comparison of different motors for EV applications [43,73,98,128].

Characteristics	IM	PMSM	SRM	DC-M
Manufacturability	5	3	4	3
Controllability	5	4	3	5
Cost	5	3	4	3.5
Robustness	5	4	4.5	3.5
Reliability	5	4	4.5	3
Lifetime	5	4	4.5	3.5
Torque ripple/noise	4.5	4	3	3.5
Technical maturity	4.5	4	3.5	5
Efficiency	4	5	4.5	3
Size and weight	4	4.5	4	3
Power Density	4	5	3.5	3
Overload capability	4	4.5	4	3
Speed range	4	5	5	2.5
Torque density	3.5	5	4	3
Potential for improvement	3	4.5	5	2.5
Trend	2	5	3	1
Total Score	67.5	68.5	64	51

compares all motor key characteristics in order to provide guidelines for choosing an appropriate motor for EV applications. PMSM is the most demanding motor, followed by IM and SRM, whereas DC motors are employed for small applications and are not chosen for EVs. Different traction motors employed in commercially available EVs and HEVs are shown in

Table 16. Motor employed in EV or HEV [113,129,130,131,132,133].

EV Model	Motor Employed	Power kW	Year
Rivian R2	PMSM	220	2026
Toyota Land Cruiser Se Volkswagen ID. GTI	PMSM PMSM	400 210	2026 2026
Lucid Gravity	PMSM	600	2025
BYD Seal	PMSM	150	2025
Tesla Model Y long range RWD BYD Seagull Tesla Model 3 Standard Range	IPM-SynRM PMSM PMSM	250 55 239	2024 2024 2023
Tesla Model Y long range RWD BYD Seagull Tesla Model 3 Standard Range	IPM-SynRM PMSM PMSM	250 55 239	2024 2024 2023
Renault Megane E-Tech	PMSM	96	2022
Porsche Taycan ST	PMSM	300	2022
Mercedes-Benz EQS	PMSM	265	2022
Kia Niro EV	IM	150	2022
Tesla Model Y SR	IM	150	2022
Tesla Model 3 long range	IPM-SynRM	258	2021
BYD Tang	PMSM	380	2021
Hyundai Ioniq 5	PMSM	125	2021
Tata Nexon	PM	85	2020
Volkswagen id.3	PM	100	2020
Kia e-Soul	PM	100	2020
Mercedes EQC	IM + IM	300	2019
Smart EQ Fortwo	PM	60	2019
BYD S2	PM	70	2019

.

4. Control Techniques of Traction Motor

There are various techniques for controlling motor drive and all have their particular significance when dealing with motor control with respect to specific applications. Some are simple but less accurate (e.g., scaler control), while others are sophisticated but produce accurate findings (DTC, FOC, MPC, etc.). As a result, traction motor control is critical while heading towards smart applications [134,135,136]. Both high dynamic response and accurate speed control are the key characteristics required from the traction drive operation. Advanced motor controllers are responsible for providing not only good steady state, but also better transient response. In general, key characteristics required from the traction motor drive include the capability of high torque at low speed for starting and climbing, high power for cruising, fast torque response, and high-power density. Moreover, it should maintain high efficiency across wide speed ranges, while being compact, lightweight, cost-effective, reliable, and robust under varying vehicle operating conditions. Figure 17 illustrates the typical speed–torque characteristic curve essential for EV and HEV applications.

4.1. Scalar Control

This is a simple speed control technique that employs voltage and frequency to offer both open and closed loop control of an induction motor. The V/f control approach is employed to increase the dynamic response and operational performance. Scalar control has a simple design, a simple structure, minimal steady-state error, and reduced cost while providing soft starting. V/f control provides improved speed and torque control for IM applications [137]. The search control is executed using small-magnitude triangle extra voltage injection. The deployment of a system that searches for the lowest possible stator current improves the efficiency of an induction motor [138].

4.2. Vector Control

The vector control (VC) method provides high-performance induction motor control by magnitude and phase. For generating torque and flux, VC relies on the orthogonal Clarke and Park transformations. It provides high starting torque, the same as separately excited DC motors, by employing two independent orthogonal variables, field current and armature current [43]. It encountered difficulty and complexity when using IM due to the electromagnetic torque and flux relationship. This problem has been fixed in the proposed optimal control of combining FOC and DTC [139].

Indirect vector control (IVC), also known as FOC, is a well-known vector control approach [140]. The FOC provides better dynamic performance as well as independent control of rotor flux and torque [44]. The orientation angle that allows perfect decoupling of the two factors is the rotor flux angle. It is equipped with PI controllers to deal with the magnitude and phase of current, torque and speed. In FOC, the modulation approach is applied. Switching vector (SV) PWM and FOC are frequently employed [141]. FOC cascades PI controllers using stator current; the error between currents is employed for better control and speed estimation. As a result, it necessitates particularly demanding operating conditions [125]. IVC of induction motor traction drive is illustrated in Figure 18.

For PMSM, VC is widely adopted due to the inherent sinusoidal back-EMF and absence of rotor current dynamics. The control objective simplifies to regulating the d-axis current and q-axis current, where the optimal operation often enforces zero d-axis current for surface-mounted PMSMs. This leads to high efficiency and fast dynamic response. However, in interior PMSMs, reluctance torque introduces an additional degree of freedom, requiring advanced control strategies such as maximum torque per ampere (MTPA) and field-weakening operation [142].

In BLDC motors, the application of conventional VC is less straightforward due to the trapezoidal back-EMF profile. Instead, quasi-vector control or modified FOC approaches are employed to approximate sinusoidal current excitation, thereby reducing torque ripple and acoustic noise. The main challenge lies in accurate commutation and current shaping, especially at low speeds where position estimation becomes less reliable [142]. Recent improvements include hybrid control schemes that combine FOC with predictive current control to enhance dynamic performance.

For SRMs, the absence of permanent magnets and the highly nonlinear inductance profile complicate the direct application of classical VC. Nevertheless, extended vector control frameworks have been proposed by defining pseudo d-q-axes or utilizing flux-linkage-based transformations. These approaches aim to achieve smoother torque production and reduced ripple. However, strong magnetic saturation and phase coupling effects limit the accuracy of such models [142]. Consequently, adaptive and nonlinear control techniques are increasingly integrated with VC to address these limitations.

Despite its advantages, VC faces several challenges across all machine types, including sensitivity to parameter variations, reliance on accurate rotor position information, and increased computational complexity. These issues become more pronounced in EV applications where wide speed ranges and varying load conditions are common.

There are several improved FOC controllers available to provide the required operational performance for EVs. EV acceleration can be improved by adjusting the maximum torque and ampere ratio with FOC. Field weakening design and FOC can be employed to increase the full speed operation of EVs. It will increase the motor’s speed range. Precision in output torque and efficiency are dependent on accurate rotor flux estimation, which is accomplished via FOC [143,144]. The FOC scheme associated issues and some suggested solutions are summarized in

Table 17. FOC issues’ proposed solution and outcomes [130,132,133,145].

Issues	Proposed Techniques/Method	Advantages	Disadvantages
Assessment of suitability	IFOC for asynchronous traction motor drives for EVs	Suitability of IFOC is verified.	No additional improvement suggested.
Degradation of speed performance due to thermal effects	LPV controller–observer is employed	Low current and supply voltage are required. Noise rejection improved. Better tracking performance.	More complicated than the conventional scheme.
Motor parameter variation impact on the performance.	MRAS estimator based on back EMF is proposed	Robust against inverter nonlinearity and control mechanism. Independent of stator inductance and resistance.	The results presented consider only 50% change in the stator parameter.
Comparison of conventional PI and Fuzzy logic controller.	Use of the ECE-15 drive cycle	Improved speed performance. Tracking performance, energy consumption and recovery.	Improvement in IFOC is not considered.
Analysis of drivetrain efficiency.	FOC PMSM. Three-phase inverters, battery connected with a DC-DC converter.	Improved power factor and efficiency. Reduced voltage distortion.	The energy optimization technique is not considered a variation in motor flux.

.

4.3. Direct Torque Control

DTC needs a switching table to determine the current state of the inverter in order to generate the correct output torque [146]. It has a PI speed controller and two hysteresis controllers. It employs two different controllers, achieving a better dynamic response. A three-level comparator serves as a torque hysteresis controller, while a two-level comparator serves as a flux hysteresis controller with a different bandwidth [89,147].

The flux hysteresis bandwidth influences motor current distortion, while the torque hysteresis bandwidth influences switching losses and switching frequency. As a result, it is critical to select the hysteresis bandwidth with care. DTC allows for precise and rapid control of IM torque and flux without the use of sophisticated controlling algorithms [148]. The use of DTC for IM traction drive control is demonstrated in Figure 19.

DTC includes switching vector modulation (SVM), switching table-based, and direct self-control. Its qualities of not being parameter dependent make it particularly significant for electric car applications, with resilience strengthened as it is not dependent on parameter variation and a smaller number of controllers. DTC is appropriate for high-speed operation and enables frequent starting/stopping and acceleration. DTC allows for the implementation of strong flux weakening control as well as dynamic motor operating. DTC has a slow start-up response, with high current and torque ripples [94,149]. However, low speed flux, torque estimate, torque control, torque ripples, and high current all have an impact on the DTC electric drive system. As a result, several variations or combinations are required for improved performance. DTC is implemented using a fuzzy logic controller (FLC) and a PI controller, and it is compared to a sliding mode control (SMC) scheme as demonstrated in

Table 18. DTC issues’ proposed solution and outcomes [131,144,145,152].

Issues	Proposed Method	Advantages	Disadvantages
Minimization of issues in conventional DTC	Utilization of a multi-level neural network instead of a conventional switching table	Torque and current ripples were reduced significantly	Efficiency improvement is not investigated. A simulation study on the common drive cycle has not been done.
Efficiency improvement of DTC	Variable flux reference selection	Efficiency improved	A very small-scale model based on simulation is employed but without using recognized drive cycles.
Control method for a four-in-wheel EV drive	An additional electronic differential system provides speed references to in-wheel motors.	Handling improved. Good dynamic and steady-state speed tracking.	High torque ripple is observed
Reduction in torque ripple in CDTC	Modified torque hysteresis controller. The MRAS estimator is employed for rotor speed and stator resistance.	Torque error, torque ripple, current THD and flux error improved.	Optimal bandwidth is employed, which may be difficult to achieve in large-scale systems for EVs.
Improved integral time- weighted absolute error (ITAE) and speed response	Fractional order PI controller.	ITEA is reduced as compared to the PI controller.	Issues associated with CDTC have still not improved.
Efficiency improvement in DTC	Stator current minimization technique. DC link voltage is employed.	Lower energy requirement. Improved efficiency. Excessive computational resources are not required. No convergence issues. Not sensitive to parameter variation.	Maximum torque/ampere is allowed during operation but below base speed.

[150]. Ben Salem et al. proposed an inventive adaptive fractional order sliding mode (FO-SM) control method specifically designed for DTC-SVM applied to IM drives, which addressed the issues of parameter variation in load and temperature with better control and reduced torque ripples [151].

In PMSMs, DTC has been adapted to exploit the fast torque response while addressing the inherent torque and flux ripple associated with hysteresis-based switching. Unlike induction motors, PMSMs exhibit a strong coupling between flux and torque due to the presence of permanent magnets. This necessitates refined switching table designs and the incorporation of SVM to achieve constant switching frequency and reduced ripple [142]. Recent advancements include model predictive DTC (MPDTC), which enhances control accuracy by predicting system behavior over a finite horizon [142].

For BLDC motors, DTC implementation is comparatively simpler due to the discrete nature of back-EMF and commutation. However, conventional DTC strategies may lead to significant torque ripple because of the non-sinusoidal characteristics. Modified DTC schemes incorporating duty ratio control and advanced switching strategies have been proposed to mitigate these effects [142]. Additionally, the integration of intelligent controllers has improved robustness under parameter uncertainties.

In the case of SRMs, DTC is particularly attractive due to the direct relationship between phase inductance and torque production. Unlike PMSMs and BLDC motors, SRMs do not require rotor flux estimation, simplifying control design. However, the highly nonlinear magnetic characteristics and significant torque ripple present major challenges. Advanced DTC approaches, including torque-sharing functions and predictive control, have been developed to address these issues. These methods aim to distribute torque production smoothly across phases, thereby reducing vibration and acoustic noise [142].

Although DTC provides fast transient response and reduced dependence on machine parameters, it suffers from variable switching frequency, high torque ripple, and difficulties in low-speed operation. These limitations necessitate the use of enhanced modulation techniques and intelligent control algorithms.

The applicability of VC and DTC across different traction motors highlights a fundamental trade-off between control precision and implementation complexity. VC offers superior steady-state accuracy and low torque ripple, making it suitable for applications requiring smooth operation, such as PMSM-based EV drives. In contrast, DTC provides faster dynamic response and simpler structure but requires additional modifications to achieve comparable performance.

Recent research trends indicate a convergence of these techniques through hybrid control strategies. For instance, predictive control frameworks are being integrated with both VC and DTC to improve dynamic performance while maintaining robustness. Similarly, AI-based approaches are increasingly employed for parameter estimation, fault diagnosis, and adaptive control across all motor types [142].

Another important direction involves the development of sensor-less control methods to eliminate mechanical sensors, thereby reducing cost and improving system reliability. This is particularly relevant for EV applications where harsh operating conditions can affect sensor performance.

5. Artificial Intelligence Controllers

Outstanding operational performance of the present artificial intelligence controllers (AICs) has piqued the interest of many researchers in recent years [144]. The primary goals of ICs are to obtain high dynamic performance while reducing the processing time and increasing the robustness against the machine parameter dependency [145,146]. Motor speed and torque are regulated using optimized parameter settings for PD, PI, and PID controllers and model predictive control (MPC) enables fault-tolerant operation of the system. A hierarchical control method integrating MPC with a torque vectoring algorithm has been proposed to reduce torque requirements and achieve significantly improved results [153]. Furthermore, advanced controllers are necessary to achieve increased accuracy and precision control. artificial neural network controller (ANNC), adaptive neural fuzzy inference system controller (ANFIS Controller), and fuzzy logic controller (FLC) are some of the advanced controllers [147,148].

5.1. MPC Controller

In a recent study, a major shift was observed towards the MPC controller to deal with the EV control system and real-time parameter improvement. ANFIS-based MPC is proposed to reduce current and torque ripples [154]. A hierarchical predictive optimal control is proposed to control IPMSM drives to improve the torque control with improved energy efficiency of 37% [153]. Hybrid MPC is employed to deal with stability control for a distributed drive vehicle on real-time conditions of different road surfaces at high speeds [155]. Another improved drive system for five-phase PMSM is proposed based on MPC. This article proposed an emerging speed tracking technique to deal with the faults in open-phase motor [156].

5.2. ANN Controller

To increase efficiency and fast processing, conventional technologies are progressing towards intelligent systems. An intelligent controller is necessary to achieve higher precision and eliminate losses in the conventional controllers. Figure 20 shows the Basic 3-inputs and 1-output structure of an ANN. Motor parameter dependency, torque ripples and efficiency are the issues in motor controllers. An adaptive neural network-based torque controller is designed for the inner current loop to mitigate torque ripple and improve propulsion motor efficiency, which improves EV performance [157,158]. An ANN-based model predictive controller is used for optimal torque performance [153]. Another ANN-based model with MPC is employed for better speed estimation [156]. ANFIS is compared to FLC; the results show that ANFIS has greater dynamic performance [159,160].

ANN is utilized as an intelligent controller to estimate motor parameters and detect motor faults [161,162]. A more robust controller is proposed in [163], in which PMSM will benefit from two innovative deep neural network (DNN)-based vector controllers. The numerical burden and complex computation are the issues to be addressed in this type of controller.

5.3. ANFIS Controller

The ANFIS controller is a hybrid of artificial neural networks and fuzzy logic. It is designed to take advantage of the unique qualities of both networks. It is an intelligent controller employed to estimate motor parameters without the use of a mathematical model. The ANFIS model prediction current control method is employed to reduce torque and current ripples in PMSM [159]. ANFIS has the problem of requiring a big data set for data training and learning, which takes a long time and requires complicated computing processing. As compared to a traditional controller, advanced ANFIS and FLC produced efficient and better HEV performance [162,163,164]. An ANFIS-based fractional order PID controller exhibits robustness against external disturbances and provides potential EV speed regulation control [165]. PI-ANFIS and particle swarm optimization (PSO)-PI-ANFIS algorithms were contrasted, and PSO-PI-ANFIS forces the motor speed to closely follow its reference. Furthermore, it models with simpler structures, which are more reliable than PI-ANFIS models [166]. The advantages of these approaches are their high accuracy and extremely fast computational speed for prediction and adjustment of the classical PI controller parameters; however, they have a complex structure [167].

5.4. FLC

Fuzzy Logic Controllers (FLCs) are intelligent controllers that use a fuzzy logic network. FLC operates in four basic steps. Fuzzification comes first, followed by the rule base, inference engine, and defuzzification. The initial stage is to determine the number of inputs, the position of fuzzy MFs, and the outputs. As stated in the equation, the inputs are error and error change [168].

$$e\left(t\right)=x_n^{*}-x_n$$

(2)

$$de\left(t\right)=e\left(t\right)-e\left(t-1\right)$$

(3)

By analyzing each input in a group, the fuzzification design characterizes the right linguistic value with inputs. This is known as a one-of-a-kind MF label. The number of MFs employed determines the precision of the FLC. Triangular or trapezoidal MFs are employed to depict the output. The architecture of an inference engine indicates the control rules and language concepts of decision-making. It is usually divided into two categories, Mamdani and Takagi-Sugeno. Due to its basic structure and easy design, the Mamdani approach is widely employed. As a fuzzy rule, the if-then clause is employed. The basic architecture of FLC is demonstrated in Figure 21.

The final stage is defuzzification, which generates the output values as crisp values. This method regulates the crispness of output MFs and makes minor adjustments. FLC has found application in the operation of HEVs as it is simple, effective, and offers the option of real-time supervisory control based on a deterministic rule-based approach [145]. When the FLC’s speed tracking performance was compared to that of the PI controller, the FLC performed better at constant speed [169]. FLC is employed for speed improvement; therefore, the speed of the motor is improved, and the error is reduced [170]. FLC can enhance starting current amplitude and conserve more power [171]. Improved adaptive FLC-based techniques are used for traction drive and energy management systems [172]. Moreover, the Neuro-fuzzy controller is more advanced and is based on a nonlinear and adaptive method. As a result of its precision, extensive supervisory control, and low error rate, this adaptable system can be employed in even more complex real-time circumstances [173]. The comparison table of AI controllers is shown in

Table 19. Comparison of AI controllers [166,174,175,176,177,178].

Aspects	MPC	ANN	ANFIS	FLC
Concept	Model-based optimization using a prediction horizon	Data-driven learning to approximate nonlinear system behavior	A hybrid of fuzzy logic + neural networks for adaptive rule-based learning	Rule-based control using linguistic variables and expert knowledge
Advantages	Handles multiple constraints Optimal control performance Fast dynamic response Suitable for multi-motor coordination	High adaptability Strong nonlinear approximation Fault tolerance Self-learning capability	Combines interpretability + learning Effective for nonlinear systems Adaptive parameter tuning More robust than pure FLC	Simple implementation No mathematical model required Robust to uncertainties Low computational cost
Limitations	High computational complexity Requires an accurate model Difficult real-time implementation at high speed	Needs large training data Lacks interpretability Generalization issues No guaranteed stability	Rule explosion problem High computational cost Complex training Scalability issues	Heuristic design No optimality guarantee Limited adaptability Performance depends on rule tuning
Applications	Torque control in PMSM drives Multi-motor coordination Stability and energy optimization	Fault diagnosis and tolerant control Parameter estimation Adaptive torque control	Adaptive speed control Hybrid EV control systems Nonlinear drive control	Low-cost EV controllers Basic speed and torque control Real-time embedded systems

.

6. Present Challenges and Recommendations

In spite of substantial research and development work, EVs and HEVs still face numerous technical and large-scale commercial adoption challenges that require more research, innovation, and technological refinement for optimized performance and mature technology.

6.1. EV and HEV Powertrain Architectures

6.1.1. Present Challenges

• High System production cost, weight optimization and complexity due to multi-source systems, such as engine, battery, power converters and generator integration, are still persistent issues.
• Limited energy density of current batteries results in restricted driving range and frequent recharging requirements.
• Limited battery capacity, cost-effective energy storage and charging infrastructure limitations remain key challenges for pure EVs.
• Energy management and packaging constraints in compact vehicle designs.
• Limited energy storage capability and inefficient energy management under highly dynamic driving conditions.
• Reliability and durability concerns in commercial fleets with high utilization rates.

6.1.2. Recommendations

• Development of an optimized lightweight design of modular and scalable architectures approach, which integrates renewable energy sources and is easier to implement both on light-duty and heavy-duty vehicles.
• Integration of advanced solid-state batteries, supercapacitors, and hybrid energy storage systems for improved driving range and lifecycle.
• Application of predictive energy management strategies using artificial intelligence.
• Enhanced energy management strategies using predictive and data-driven models can further optimize fuel economy and reduce emissions in hybrid power-split configurations.
• Research on power electronics with wide-bandgap semiconductors (SiC, GaN) to improve efficiency and reduce system losses.

6.2. Traction Motors

6.2.1. Present Challenges

• Dependence on rare-earth permanent magnets for PMSMs increases the overall system cost and supply chain vulnerability.
• IMs suffer from relatively high losses at low load and limited efficiency at high speeds.
• BLDC motors and SRMs offer potential alternatives but require further refinement in noise reduction, efficiency, and control mechanism complexity.
• Thermal limitations reduce lifespan and reliability in high-power applications.
• Packaging and weight constraints limit the adoption of large machines in compact EV design.

6.2.2. Recommendations

• Advancement and development of rare-earth-free designs such as SRMs and advanced induction motor topologies.
• Exploration of axial flux and transverse flux machines offering higher torque density and compact design.
• Improved cooling strategies, such as direct liquid cooling and advanced thermal interface materials, for better heat dissipation.
• Design of fault-tolerant motor architectures to ensure reliability in safety-critical applications.
• Optimization of multi-motor propulsion systems (e.g., in-wheel motors) for distributed drive and enhanced control flexibility.
• Integration of novel materials, additive manufacturing, and fault-tolerant designs is expected to play a key role

6.3. Motor Control Techniques

6.3.1. Present Challenges

• Conventional methods such as vector control and DTC are sensitive to parameter variations during operation and disturbances.
• Both aforementioned methods are employed for high-performance applications such as in EVs and HEVs, but suffer under dynamic and uncertain driving conditions, torque ripple and high efficiency across wide operating ranges.
• High computational demands in real-time control for nonlinear and multi-variable systems are required.
• Face challenges in ensuring robustness under variable driving conditions, limited robustness in handling motor faults, inverter faults, and grid interaction scenarios.
• Cybersecurity risks in intelligent and connected control systems are becoming increasingly important for safety and reliability in autonomous and connected EVs.

6.3.2. Recommendations

• Incorporation of artificial intelligence, machine learning, and adaptive control for predictive torque control, less computational burden and efficiency optimization.
• Development of self-tuning algorithms that can adjust controller parameters automatically under changing load and environmental conditions.
• Advanced sensor-less control techniques are also important to reduce hardware costs and improve reliability.
• Integration of fault-diagnosis and fault-tolerant control techniques for enhanced safety.
• Research into distributed and cooperative control strategies for multi-motor EV architectures.
• Enabling motor controllers to support vehicle-to-grid (V2G) operations, including bidirectional power flow and grid stabilization.

These challenges in EV and HEV architectures, traction motors, and motor control require interdisciplinary collaboration across materials science, power electronics, control theory, artificial intelligence and system integration. Future EV and HEV systems will likely rely on high-efficiency, rare-earth-free traction motors, modular powertrain designs, and intelligent controllers capable of adaptive learning. Together, these developments will accelerate the transition toward sustainable, reliable, and cost-effective electrified transportation.

7. Conclusions

This paper has presented a structured and critical assessment of electric and hybrid electric vehicle powertrain architectures, traction motor technologies, and advanced control methodologies. A structured comparative framework was employed to evaluate series, parallel, and series–parallel architectures with respect to efficiency, cost-performance balance, structural complexity, and commercial feasibility. The review further analyzed dominant traction motor technologies with emphasis on torque–speed capability, efficiency mapping and material considerations, identifying both technological breakthroughs and unresolved limitations. Although PMSM, IM, and related topologies demonstrate high performance, challenges associated with rare-earth utilization and scalability remain open research concerns. Advanced control strategies, including vector-based methods, direct torque approaches, model predictive control and emerging AI-enabled frameworks, were comparatively evaluated in terms of dynamic response, robustness, computational demand, and scalability. Key research gaps were explicitly identified in areas such as rare-earth material dependency, high-speed efficiency optimization, integrated energy management, and real-time intelligent control implementation. The findings underscore the need for multidisciplinary innovation in next-generation motor design, power electronics integration, and adaptive control architectures to achieve reliable, high-efficiency, and economically sustainable electrified mobility solutions.