Design of Single-Stage Management System for Grid-Connected Photovoltaic Sustainable Power Generation and Its HVRT Technology with Energy Storage Coordination

Abstract

With the rapid development of sustainable photovoltaic power generation, energy-storage-coordinated grid-connected photovoltaic systems have been widely adopted to stabilize power output and enhance grid adaptability. Aiming at the low fault tolerance of conventional photovoltaic grid-connected systems under grid voltage swell disturbances, this paper designs a single-stage power management system for grid-connected photovoltaic generation and studies its energy-storage-coordinated high-voltage ride-through (HVRT) technology. The single-stage topology boasts simple structure, low cost and high conversion efficiency, yet faces prominent stability risks under voltage swell faults. The system integrates photovoltaic units, energy storage modules and grid-connected interfaces to implement flexible bidirectional power dispatching. A three-phase AC/DC converter realizes photovoltaic maximum power point tracking (MPPT), and the energy storage module connects to the DC bus via a dual half-bridge (DHB) converter to restrain power fluctuations. Under HVRT faults, the energy storage coordination strategy elevates DC bus voltage to maintain stable grid-tied operation without disconnection. Different from schemes requiring extra hardware or complicated control optimization, the proposed method realizes stable bus voltage regulation and flexible energy scheduling with zero additional hardware cost. Simulations and experiments validate the rationality, feasibility and outstanding fault-ride-through performance of the designed system.

1. Introduction

Global industrial expansion and ever-rising energy consumption have accelerated the depletion of fossil-fuel resources, accompanied by excessive carbon emissions and severe ecological degradation, which drives profound transformations within worldwide energy frameworks [1,2,3]. To achieve a balance among energy security targets, socioeconomic advancement and ecological conservation, nations across the globe are devoting substantial efforts to developing low-carbon energy infrastructures and advancing renewable-energy industries. Statistics released in the 2026 Global Energy Report by the International Energy Agency reveal that renewables contribute nearly 60% of newly-added global energy demand. Photovoltaic generation, characterized by low-carbon properties, abundant resource reserves and flexible deployment conditions, has become the fastest-expanding clean energy sector and reshaped conventional energy supply–demand patterns. Currently, global energy development has entered a brand-new stage, where fossil fuels are gradually phased out, and new energy sources dominate capacity increments. Constructing modern clean energy systems centered on photovoltaic and wind power has turned into a prevailing consensus for sustainable energy development [4].

Figure 1 visualizes global renewable-energy development forecasts from the International Renewable Energy Agency. Figure 1 shows global electricity-generation outcomes, quantitatively indicating that renewables will account for 86% of total global power generation by 2050, among which solar PV and wind power occupy the dominant share of renewable output. Figure 1 also presents cumulative installed-capacity projections; the total installed capacity of renewables will approach 19,000 GW in 2050, with solar PV reaching 8519 GW and wind power 6044 GW. Such massive penetration of photovoltaic generation imposes stringent requirements on the fault-ride-through performance of grid-connected photovoltaic equipment.

Photovoltaic power stations are commonly integrated with energy storage devices to form hybrid photovoltaic–energy storage systems, which are divided into grid-connected and off-grid topologies, as illustrated in Figure 2. Figure 2a demonstrates the grid-connected PV–storage configuration: the photovoltaic array interfaces with the public utility grid via an inverter, an electricity meter realizes grid-energy metering, and local loads are jointly supplied by PV output and grid power. Operating in grid-tied status, this system must satisfy mandatory grid-code specifications, including high-voltage ride-through capability under grid voltage disturbances. Figure 2b shows the off-grid PV–storage system, which operates autonomously without public-grid support. A charge controller and energy storage terminals are essential for bus voltage stabilization, and all local loads are fully supplied by the PV–storage combination. Compared with off-grid systems, in practical grid-tied scenarios, grid voltage swell faults impose strict requirements on system fault-ride-through capability, among which high-voltage ride-through (HVRT) is one of the key technical indicators to guarantee continuous grid-connected operation [5,6]. Among various system topologies, single-stage grid-connected PV systems have attracted significant attention due to their advantages of low cost, compact size, and high conversion efficiency [7,8]. However, the inherent intermittency of PV power poses significant challenges to grid stability, especially under fluctuating irradiance and load conditions [9,10].

To address the issue of PV intermittency, integrating energy storage devices into PV systems has become an effective solution. Multi-port power converters, serving as key interfaces for flexible integration of an energy storage system (ESS), have been extensively studied. Reference [11] proposed a four-port power converter that integrates energy storage with distributed generation, thereby improving power quality. For high-step-up applications, references [12,13] introduced a three-port converter combining PV and energy storage, along with a control strategy designed to maintain MPPT under varying irradiance and load conditions, enabling efficient multi-mode operation. Recent review papers have systematically classified multi-port converter topologies and their control strategies in renewable energy integration applications [14,15]. Nevertheless, these studies primarily focus on energy management and efficiency optimization under normal operating conditions, with insufficient attention paid to the system’s ability to ride through grid faults. With the rapid penetration of renewable energy, various hybrid energy storage configurations are widely adopted to cope with power fluctuation of photovoltaic generation. Hydrogen-battery hybrid storage provides a feasible solution for stable renewable power utilization, which has been modeled and analyzed in existing studies [16,17]. Apart from performance optimization, safety is another critical concern for large-scale lithium-ion battery energy storage deployed at grid side [18]. Benefiting from continuous breakthroughs in material and packaging techniques, battery technologies are evolving rapidly to satisfy the requirements of future grid-tied energy storage scenarios [19]. Inspired by the above progress, this paper investigates a three-port single-stage PV–ESS system with coordinated HVRT capability.

Beyond normal operation, fault-ride-through capability is equally critical for grid-connected PV systems during grid disturbances. Currently, low-voltage ride-through (LVRT) strategies for PV systems have been widely investigated [20,21,22]. In contrast, HVRT technology remains in its early stages [23,24,25]. Grid voltage surges caused by large-scale load shedding or grid faults may lead to power backflow through the grid-side converter and overvoltage on the DC bus, potentially causing system disconnection. To address this challenge, several HVRT control strategies have been proposed. Reference [26] studied the impact of dynamic voltage support (K-factor) and active power recovery rate on power swing blocking protection during LVRT/HVRT processes in large-scale PV plants, revealing the evolution patterns of system impedance trajectories during fault ride-through. Reference [21] proposed an inverter control technique capable of achieving both LVRT and HVRT simultaneously by adjusting terminal voltage to meet grid voltage limits and utilizing residual inverter capacity to provide active power support. Reference [27] presented an improved HVRT control strategy incorporating variable DC bus voltage reference and dynamic reactive power support to mitigate overmodulation issues caused by voltage surges. Reference [17] applied model predictive control to a single-stage dual DC-port inverter, expanding candidate voltage vectors via discrete space vector modulation to achieve grid current tracking while addressing multiple control objectives. Reference [28] proposed a hybrid power decoupling method for a railway converter based on a through-type single-phase inverter, resolving strong coupling between active and reactive power among three ports. However, these methods typically require additional hardware or complex control modifications, and systematic research on HVRT implementation within multi-port converter architectures remains limited.

Recent advances further explore multi-port converter topologies and energy-storage-assisted fault mitigation for PV applications [3,8,22]. Nevertheless, most existing three-port PV–ESS converters focus on steady-state energy management and efficiency optimization, without dedicated HVRT inner-loop design for grid voltage swell events [14]. Meanwhile, updated grid-code specifications impose stricter overvoltage magnitude and fault-duration constraints for inverter-based resources, which raise higher requirements for the fault tolerance of multi-port power interfaces [10]. However, these methods typically require additional hardware or complex control modifications, and systematic research on HVRT implementation within multi-port converter architectures remains limited.

Building upon the above studies, this paper proposes a single-stage PV–storage grid-connected system based on a three-port converter with HVRT capability. In the proposed system, PV maximum power tracking is always controlled by the three-phase AC/DC converter. The energy storage device is flexibly integrated via the converter, dynamically suppressing power fluctuations under different operating conditions and thus enhancing overall system efficiency. Under high-voltage grid faults, the energy storage device actively raises the DC bus voltage to maintain system stability, enabling continuous operation without disconnection throughout the entire fault duration until the grid returns to normal. Simulation and experimental results verify the effectiveness of the proposed system.

The remainder of this paper is organized as follows: Section 2 introduces the system architecture and operating principle of the proposed three-port converter; Section 3 provides a detailed analysis of the system’s power flow characteristics and HVRT control strategy; Section 4 validates the effectiveness of the proposed system and control method through simulation and experimental results; Section 5 concludes the paper.

2. PV–ESS System Structure and Energy Flow

2.1. Single-Stage PV–ESS System Structure and DHB Reuse Half-Bridge Modulation Principle

The basic framework of the PV–storage grid-connected system is illustrated in Figure 3, where solar PV units and energy storage (ES) are connected to the power management system. The power management system realizes coordinated power dispatch between photovoltaic sources, energy storage units and the utility grid, supporting bidirectional energy flow among multiple ports. The bidirectional arrow between the power management system and the grid indicates that energy can flow in both directions, from the system to the grid when excess PV power is fed into the grid, and from the grid to the system when PV and storage are insufficient to meet the load demand.

Figure 4 shows the topological structure of the three-port photovoltaic–energy storage grid-connected system proposed in this paper. The system includes two DC buses, Bus1 and Bus2, which can exchange energy bidirectionally through a non-isolated bidirectional DHB DC converter, and each supplies power to its respective connected loads.

Port1 is the DC bus voltage input port, Port2 connects to the energy storage device for smoothing fluctuations in photovoltaic output, and Port3 is dedicated to HVRT, enabling the system to meet continuous operation requirements without disconnection by increasing the DC-side voltage when grid voltage suddenly rises.

The photovoltaic module achieves MPPT via a three-phase converter, which adjusts the DC bus voltage to continuously follow the reference voltage corresponding to the maximum power point of the PV array, thereby forming a single-stage grid-connected structure. Compared with traditional two-stage configurations, this topology eliminates the intermediate DC/DC boost stage, improving energy transmission efficiency while reducing system size and cost. In this work, the term “single-stage” indicates that the photovoltaic port, energy storage port and grid-tied port share a common dual-half-bridge conversion unit. No additional cascaded DC–DC or DC–AC conversion modules are inserted among these ports, even though the dual-half-bridge topology performs equivalent buck-boost-type modulation functions.

For a conventional non-isolated bidirectional DC converter combined with a DHB system, six switches are typically required: two switches on one side form a Buck/Boost circuit to adjust the voltage level on the energy storage side, while the other four switches are dedicated to enabling bidirectional power transfer in the DHB. However, in this proposed topology, the half-bridge switches S_i3 and S_i4 on the primary side of the DHB converter exhibit functional reuse—these two switches alone can simultaneously fulfill both tasks. This eliminates the need for two additional switches compared to conventional designs, thereby reducing hardware cost and control complexity without compromising system functionality.

Under normal operation, this shared bridge arm must simultaneously support two operating modes. On one hand, it participates in the standard phase-shift modulation of the DHB converter, forming the DHB together with the secondary-side bridge arm composed of S_i5 and S_i6. The system employs a single-phase-shift strategy, where the power transfer direction and magnitude between the two DC buses are controlled by adjusting the phase difference ϕ between the driving signals of the primary and secondary bridge arms. When ϕ > 0, power flows from Bus1 to Bus2, charging the energy storage; when ϕ < 0, power flows in reverse, discharging the energy storage. On the other hand, switches S_i3 and S_i4, together with the filter inductor L₂ on the energy storage side, also form a basic Buck/Boost circuit. When energy needs to be stepped up from the low-voltage energy storage side to the high-voltage DC bus, the system adjusts the duty cycle D of the lower switch S_i4 to satisfy the relationship V_Bus2 = V_ESS/(1 − D). Conversely, when energy is drawn from the high-voltage bus to charge the energy storage at a lower voltage, the voltage reduction is achieved through chopper control of the upper switch S_i3. The bidirectional power transfer between the energy storage port and Port1 is jointly regulated by the phase shift angle ϕ and the duty cycle D of the shared bridge arm, with the power expression given as

$$P={\displaystyle \frac{n{V}_{\mathrm{BUS}1}{V}_{\mathrm{ESS}}\varphi (\pi -\left|\varphi \right|)}{2{\pi }^2{f}_s{L}_3(1-D)}}$$

(1)

In the equation, n is the turns ratio of the high-frequency transformer between the primary and secondary sides, f_s is the switching frequency, and L₃ is the phase-shifting inductor of the DHB. The phase angle ϕ controls the direction of power transfer; the duty cycle D adjusts the primary-side bus voltage V_Bus2 of the DHB to match the primary-side voltage of the isolated converter with the voltage at the energy storage port. In this dual half-bridge converter, the phase-shift angle ϕ and duty ratio D serve as two independent modulation degrees-of-freedom with different physical functions. The phase-shift angle ϕ mainly regulates the transmitted active power among three ports by adjusting the phase difference between bridge-arm square-wave voltages. The duty ratio D determines the effective time-averaged voltage of each bridge arm and undertakes port voltage regulation tasks.

To further clarify the operating mechanism of this DHB converter under conventional phase-shifted mode, Figure 5 shows the waveforms of the system under normal operation. As shown, the drive signals for the left-leg switches S_i3 and S_i4 are complementary, as are those for the right-leg switches S_i5 and S_i6. Meanwhile, there is a phase difference of ϕ between the drive pulses of S_i3/S_i4 and S_i5/S_i6. V_cd and V_ab represent the primary-side and secondary-side voltages, respectively, while i_L2 and i_L3 denote the currents through the energy storage inductor and the DHB phase-shifting inductor, respectively.

2.2. Energy Flow Analysis of Single-Stage PV–ESS System

Figure 6 illustrates the energy flow topology of the system. All electrical power generated by the photovoltaic array is first collected at DC Bus1, making it the central node for power coordination throughout the system, and maintaining its voltage stability is crucial for safe operation.

When a voltage surge fault occurs in the power grid, DC Bus2 works in conjunction with the auxiliary port to provide the necessary voltage support for HVRT by adjusting the voltage distribution characteristics of the bus. Power exchange among the energy storage unit, the two DC buses, and the grid is bidirectional. Both DC Bus1 and Bus2 have local loads; the energy storage device acts as a supplementary power source to balance the system when photovoltaic output is insufficient, and absorbs excess energy for storage when photovoltaic generation is abundant. On the grid side, the system operates in rectifier mode during overall power deficit to maintain stability, and switches to inverter mode when there is surplus power to feed the excess back into the grid.

The energy flow architecture demonstrates that the proposed multi-port photovoltaic–storage system possesses highly flexible power-dispatching capability. Its diverse power electronic interfaces provide reliable power path support for HVRT control, thereby enhancing the system’s operational reliability and overall energy utilization efficiency under complex operating conditions.

For DC Bus1, the power balance relationship can be summarized as: When the three-phase converter operates in rectifier mode, if P_Grid + P_PV − P_Load1 > 0, the surplus power is transmitted from the dual half-bridge converter to Bus2; conversely, when the difference is less than zero, power is supplied from Bus2 to Bus1. When the three-phase converter operates in inverter mode, if P_PV − P_Grid − P_Load1 > 0, the excess power is delivered to Bus2 via the dual half-bridge converter; conversely, if the expression is less than zero, the power flow reverses.

For DC Bus2, the power balance relationship can be described as: When the energy storage device is in discharge mode, if P_ESS − P_U3 − P_Load2 > 0, the excess power is fed to Bus1 through the dual half-bridge converter; if the difference is less than zero, power is supplemented from Bus1 side to Bus2 side. When the energy storage device is in charging mode, power is uniformly transmitted from Bus1 to Bus2 via a dual half-bridge converter.

To quantitatively illustrate the hardware characteristics of the proposed shared-half-bridge three-port topology, component-level comparison with the conventional two-stage PV–ESS topology is provided in

Table 1. Component-count comparison between conventional two-stage PV–ESS and the proposed topology.

Component Item	Conventional Two-Stage PV–ESS	Proposed Shared-Half-Bridge Three-Port Topology
Power semiconductor switches	10	8
Inductors	3	2
Capacitors	4	3
High-frequency transformer	Required	Required
Gate-driver channels	10	8
Voltage/current sensors	8	7
Auxiliary power circuits	1 set	1 set
Digital control channels	12	10
Relative hardware-cost estimation	Baseline	Slightly reduced (semiconductor-oriented)

.

As shown in Table 1, benefiting from the functional reuse of half-bridge power devices, the proposed topology decreases the number of main power switches, corresponding gate-drive circuits and partial sensors. Nevertheless, passive components, transformer and auxiliary circuits still occupy a considerable proportion of the total hardware expenditure. Therefore, the reduction of semiconductor devices does not absolutely bring overall-system cost reduction. For this reason, this paper only claims a potentially reduced semiconductor component count, instead of drawing a definite conclusion on total hardware cost.

3. PV–ESS Operating Mode and Control Structure

3.1. Single-Stage PV–ESS Grid-Connected Operation Mode

The stability of Bus1 and the MPPT of the PV module are achieved through three-phase converter control. The ESS performs charge and discharge control according to the system’s overall energy efficiency optimization requirements. The stability of Bus2 and the ESS charge/discharge operations are realized via DHB control. HVRT is accomplished by controlling the voltage at Port3 u_Port3.

Under normal operating conditions, Port3 remains inactive, meaning switch S_i2 stays closed. In this state, all PV power is delivered to the DC bus side. If the PV power is insufficient, the storage system supplies energy to the DC bus via the DHB. If the PV power exceeds demand, the surplus energy is transferred from the DC bus to the storage device through the DHB. The operational modes are illustrated in Figure 7.

The control scheme for the Bus1 voltage u_Bus1 is shown in Figure 8a. Under grid-connected operation, a dual-loop control of voltage and current is employed to maintain real-time matching between voltage and power, ensuring stable DC bus voltage. By adjusting the bus voltage, MPPT for the PV module is achieved. Under different operating conditions, the converter can operate in either inverter or rectifier mode to meet system requirements.

The flow of ESS energy is controlled by DHB, as shown in Figure 8b, using a classical phase-shift control method. The energy storage current i_ess is compared with its reference value i_ess^*, and the resulting signal is processed through phase-shifted PWM to generate duty cycle signals that control the complementary conduction of switches S_i3 and S_i4, as well as S_i5 and S_i6. A phase difference exists between the drive signals of adjacent bridge arms, enabling control over the direction of energy flow.

3.2. PV–ESS Grid-Connected High-Voltage Ride-Through Mode

When symmetrical voltage swell fault occurs at the point of common coupling, the AC-side voltage of the grid-tied inverter rises sharply. For conventional single-stage photovoltaic grid-connected topologies, the DC-link voltage will be pulled up accordingly. Once the DC bus voltage exceeds the withstand threshold of power devices, overvoltage protection will be triggered, and the inverter will be forced to disconnect from the utility grid. Large-scale tripping of distributed PV units further causes power imbalance, aggravates voltage fluctuation and endangers the security and stability of the whole power grid, which makes high-voltage ride-through an indispensable capability for grid-tied renewable energy systems according to modern grid codes.

The conceptual diagrams shown in Figure 9 illustrate the macroscopic operating behavior and grid-code requirement of HVRT. Under normal operation conditions, the PV–ESS system transfers power to the utility grid following given power dispatching commands. When grid voltage swell takes place during the fault window, the system is required to remain connected rather than trip off immediately. It should maintain continuous operation throughout the fault duration and keep tracking the maximum power point of photovoltaic arrays. After fault clearance and grid voltage returns to the allowable range, the system smoothly resumes its pre-fault normal grid-tied operation state.

To ensure that the PV array can continuously perform MPPT control during energy variations, the voltage at Port3 u_Port3 must be adjusted in real-time to maintain the PV array’s output voltage at the required level for maximum power tracking.

The HVRT operation mode is shown in Figure 10b. During the HVRT process, Port3 becomes active, and the voltage at Port1 u_Port1 equals the sum of the u_Port3 and the PV output voltage u_PV, where u_PV is determined by the MPPT.

To raise the DC bus voltage u_Port1, the u_Port3 is also increased, with the required increment being u_Port3 = u_Port1 − u_PV. HVRT is achieved by controlling the voltage at Port3. The control scheme for Port3 voltage u_Port3 is shown in Figure 10a.

The stability of Port3 voltage u_Port3 is regulated by the DC bus voltage u_Bus2. Under normal system operation, the voltage at Port3 is zero. When a high-voltage fault on the grid side is detected, Port3 begins to operate. At this point, within the overall system structure, the DC bus voltage u_Bus1 equals the sum of the Port3 voltage u_Port3 and the PV output voltage u_PV. While maintaining constant PV output, the reference value of u_Port3 becomes an incremental variable u_HVRT, increasing the voltage at Port3 so that the DC bus voltage rises to a higher level. This ensures MPPT operation of the PV module and maintains stable system performance. When the grid voltage returns to normal, u_HVRT drops to zero, Port3 stops operating, and the system reverts to its stable state.

Dynamic interactions among PV MPPT, Bus-1 voltage control, Bus-2 voltage control, ESS current control, Port-3 voltage control and grid-side current control are analyzed in this subsection. In normal operation, the PV MPPT serves as an outer loop providing PV-port voltage reference, Bus-1 and Bus-2 voltage regulators stabilize intermediate bus voltages, the ESS current loop manages charging/discharging power, and the grid-side current loop regulates grid-injected power. A bandwidth-separation scheme is utilized to alleviate cross-coupling: MPPT and outer voltage loops adopt low bandwidth, while inner current loops are configured with higher bandwidth. Once HVRT mode is activated, Port-3 voltage control is assigned higher priority to elevate the DC bus voltage. MPPT remains functional with unchanged control bandwidth, and the reference signals of bus-voltage loops and grid-current loop are dynamically coordinated to prevent controller saturation. Even though partial coupling persists during HVRT transients, the bandwidth difference decouples slow outer-loop dynamics from fast inner-current responses and mitigates mutual disturbances. Bode-plot-based stability assessment demonstrates satisfactory phase margin and gain margin under both normal and HVRT conditions.

The ZVS realization of the DHB-based converter is highly dependent on the inductor current direction, phase-shift angle distribution, and load power conditions. For each bridge device, the resonant inductor current must maintain a sufficient negative commutation margin before the switch turn-on instant to completely discharge the parallel capacitor, which is the essential prerequisite for reliable ZVS achievement. Under steady-state conditions, the phase-shift ratio determines the amplitude and direction of the circulating inductor current, thereby defining the valid ZVS operating boundary. Within medium and heavy load ranges, the reactive circulating energy is adequate to guarantee full ZVS for all main switches. With the decrease of load power, the commutation current margin gradually declines, resulting in a narrowed ZVS range under light-load conditions.

As shown in

Table 2. Topology and feature overview of typical PV–ESS grid-connected systems.

Refs.	Topology and Energy Storage Configuration	HVRT Capability	Key Limitation
[5,6]	Parallel/hybrid PV–battery grid-connected system with battery integrated with PV inverter	No	Analyzes weak grid stability and fault-ride-through requirements; lacks practical HVRT implementation scheme
[7,8,12]	Single-stage/three-port/multiport PV–ESS converter, energy storage integrated on DC-side via multiport converter	No	Focuses on normal condition energy management and power flow optimization; grid voltage swell fault is not considered
[17,28]	Dual-DC-port/special-scenario multi-port PV–ESS inverter equipped with secondary energy storage DC port	No	Optimizes modulation and special application control; no dedicated HVRT mechanism for utility grid faults
[3]	Single-stage PV grid-connected system without energy storage port	Yes	Realizes HVRT by model predictive control, but cannot smooth intermittent photovoltaic power
[27]	Single-stage PV grid-connected system without energy storage port	Yes	HVRT depends on dynamic reactive power support, lacks energy storage active power buffering capacity
This paper	Single-stage three-port PV–ESS grid-connected system with shared-half-bridge DHB; embedded ESS connected by DHB converter	Yes	Realizes energy-storage-coordinated HVRT via auxiliary Port3; no additional hardware required; implements power smoothing and fault ride-through simultaneously

, the existing PV–ESS integrated schemes [5,6,7,8,12,17,28] concentrate on energy management and normal-state power regulation without practical HVRT function. Although in the literature [3,27] achieve HVRT for single-stage PV systems, these solutions do not integrate energy storage and are unable to suppress photovoltaic power fluctuations. Compared with previous studies, the proposed three-port single-stage system integrates photovoltaic, energy storage and grid interfaces. It realizes energy-storage-coordinated HVRT without extra hardware, which fills the research gap between PV–storage integration and high-voltage fault ride-through.

Apart from saving two power semiconductor switches, the shared-half-bridge three-port architecture realizes simultaneous power smoothing and HVRT within one single-stage framework without auxiliary hardware circuits. The auxiliary Port3 reuses existing DHB control loops without additional independent controllers, so control complexity does not increase significantly. During voltage swell faults, the energy storage module actively modulates DC bus voltage rather than only relying on inverter-side reactive-power adjustment. It achieves millisecond-level dynamic adjustment and stronger tolerance for deep grid overvoltage events. The above quantitative performance comparison is summarized in

Table 3. Performance indicators comparison of different PV–ESS grid-connected schemes.

Reference	Extra Hardware	Control Complexity	Dynamic Response	Fault Tolerance
[5,6]	Yes	Medium	—	Weak
[7,8,12]	Yes	Medium-high	—	Medium
[17,28]	Yes	High	Millisecond level	Medium
[3]	No	High	12 ms	Medium
This paper	No	Low-medium	≤7 ms	Strong

.

As shown in

Table 4. Comprehensive performance comparison with typical categories of existing three-port PV–ESS and HVRT solutions.

Comparison Item	Shared-Bridge-arm Three-Port Scheme	Conventional Single-Stage HVRT Scheme	Traditional TAB Three-Port Scheme	This Work
Topology	DAB-based three-port	Single-phase full-bridge	Triple active bridge	Shared-half-bridge three-port single-stage
Semiconductor count	12	8	12	8
Inductor/Capacitor	3/4	2/3	3/4	2/3
Isolation	Yes	No	Yes	Yes
Switching frequency	20 kHz	20 kHz	20 kHz	20 kHz
Control complexity	High	Medium-high	High	Low-medium
HVRT voltage range	1.1–1.2 p.u.	1.1–1.25 p.u.	Not supported	1.1–1.3 p.u.
Response time	11 ms	12 ms	—	≤7 ms
Peak efficiency	94.3%	94.5%	94.7%	95.2%
Experimental validation	Low-voltage prototype	Low-voltage prototype	Low-voltage prototype	Low-voltage reduced-scale prototype

, compared with the existing representative scheme categories, the uniqueness of this work is reflected in topology and control. On the topological level, the proposed shared-half-bridge structure reuses power devices to construct Port3 for energy storage access, without adding extra semiconductor devices. On the control level, the Port3 HVRT mechanism does not rely on additional independent controllers. Energy-management loops for normal-condition power smoothing and HVRT fault modulation are organically coupled within one unified phase-shift and duty-cycle control framework. Under grid voltage swell faults, Port3 actively adjusts the transmitted power to modulate the DC bus voltage rather than only adopting reactive-power compensation on the grid side. This coupled control realizes energy management and fault ride-through simultaneously, avoiding mode-switching oscillation that exists in many separate-control schemes.

Although comprehensive ESS capacity sizing and economic optimization require long-term operational data and multi-objective scheduling algorithms, which are beyond the scope and page limit of this manuscript, the critical ESS operating boundaries for guaranteed HVRT performance are quantitatively supplemented in this section. Under grid voltage swell faults of 1.1–1.3 p.u., the ESS port provides transient power coordination, where the instantaneous regulating power ranges from 0.12 p.u. to 0.35 p.u. Within the standard fault duration of 0.1–0.8 s, the energy exchanged between the converter and ESS is limited, indicating that the proposed HVRT strategy does not rely on large-capacity energy storage.

Combined with the DHB modulation and ZVS constraints, the valid ESS operating voltage range for stable HVRT is determined as 0.6–1.1 p.u. rated voltage, and the maximum transient current stress under extreme swell faults is quantified as 1.15 p.u. In terms of SOC constraints, the proposed HVRT strategy maintains full fault-ride-through capability within the SOC range of 20–80%. When SOC is beyond this range, the ESS lacks sufficient transient charging or discharging margin, which shrinks the effective HVRT operating region and deteriorates dynamic regulation performance.

4. Simulation and Experimental Validation

4.1. Simulation Verification

All simulation cases in this work are implemented in MATLAB 2019B. Circuit parameters, switching frequency and device settings strictly match the parameters of the down-scaled experimental prototype described in Section 4.2, which guarantees consistency between simulation and hardware-test conditions.

Due to the variety of operating conditions, simulations were conducted under different scenarios to verify the system’s working principle. The PV source in simulation adopts a physical photovoltaic array model with realistic nonlinear I-V and P-V output characteristics. The simulation is performed under standard test conditions (STC): solar irradiance of 1000 W/m² and PV cell temperature of 25 °C. The irradiance and cell temperature remain constant throughout all simulation cases, so the maximum-power-point voltage of the PV array is fixed for MPPT performance evaluation. As shown in

Table 5. Simulation and experimental prototype parameters.

Parameter	Simulation/Experimental Prototype Value
Nominal DC Bus1 voltage uBus1	100 V
Nominal DC Bus2 voltage uBus2	50 V
Port3 adjustable voltage range	0–160 V
Rated PV output voltage uPV	100 V
ESS rated power	1.2 kW
MOSFETs	600 V/30 A
DHB high-frequency transformer turns ratio n	2:1
Switching frequency fs	20 kHz
DHB phase-shift inductor L3	45 μH
ESS filter inductor L2	120 μH
DC bus capacitors CBus1, CBus2	2200 μF
ESS valid SOC operating range	20–80%
Valid ESS port voltage range	0.6–1.1 p.u.
Outer voltage-loop bandwidth	150 Hz
Inner current-loop bandwidth	1200 Hz
MPPT sampling period	10 ms
Grid nominal line-to-line voltage	380 V (50 Hz)
HVRT test swell amplitude range	1.1–1.3 p.u.
HVRT fault duration range	0.1–0.8 s

, the simulation in this work adopts identical nominal electrical parameters as the low-voltage reduced-scale experimental prototype. Parasitic parameters including wiring inductance and device junction capacitance are neglected within the simulation model.

Case I: ESS collaboratively regulates PV grid-connected power. As shown in Figure 11, the simulation duration is 1 s.

From 0–0.5 s, the energy generated by the PV system equals the grid demand, so the energy storage unit remains inactive, and the system operates in a steady state.

From 0.5–1 s, the grid demand decreases while PV generation remains relatively high, resulting in excess energy. This surplus energy is transferred via the DHB to charge the energy storage unit. The voltage at Bus2 u_Bus2 is maintained stably at 480 V through DHB control, while the PV output remains constant, demonstrating that the system consistently achieves MPPT for the PV module. This confirms the effectiveness of the control strategy, as well as the energy storage unit’s ability to regulate PV output.

Case II: ESS collaborates to complete HVRT. As shown in Figure 12, the simulation time is 2 s.

In the HVRT simulation scenario occurring from 0.6 s to 1.4 s, the nominal grid line-to-line voltage is 380 V (50 Hz); the grid voltage swell is 1.3 p.u., corresponding to a 30% voltage rise; the fault duration is 0.8 s; the fault type is symmetrical three-phase voltage swell, and the voltage at the point of common coupling (PCC) rises to 1.3 p.u. synchronously. The rated DC bus nominal voltage is 700 V, and the transient maximum DC-link voltage is limited to 860 V according to the withstand-voltage rating of the selected power semiconductor devices. Under this fault, the terminal voltage of Port3 reaches 160 V. According to the converter power-transfer constraint, the maximum supportable HVRT overvoltage amplitude provided by Port3 is 1.35 p.u. Beyond this threshold, the system needs to cooperate with power-limiting protection logic.

Under normal grid voltage operation, the system remains in a stable state during the intervals of 0–0.6 s and 1.4–2 s. During this period, Port3 is inactive, the DC bus voltage u_Bus1 is 700 V, u_Bus2 is 400 V, and the voltage at Port3 u_Port3 is 0 V, with both PV output and grid voltage maintained at steady-state levels. Between 0.6 and 1.4 s, a high-voltage fault is detected on the grid, triggering Port3 to activate.

During the interval of 0.6–1.4 s, as the grid voltage rises, the Buck/Boost converter controls the increase in Port3 voltage u_Port3, raising it to 160 V. Simultaneously, the DC bus voltage increases to adapt to the elevated voltage level caused by the high-voltage fault—specifically, u_Bus1 rises to 860 V—enabling the system to continue operating continuously without disconnection until the fault clears. Throughout this process, the PV system maintains MPPT.

Simulation results demonstrate that the system achieves MPPT for PV modules through coordinated control between ports and buses. Meanwhile, the DHB can adjust energy flow direction according to system requirements. During grid overvoltage faults, the energy storage unit flexibly switches its connection mode, regulating the DC bus voltage distribution via control of Port3 voltage u_Port3 to elevate the DC-side bus voltage level, thereby enabling HVRT and maintaining MPPT for PV modules. As a result, the system remains stable and continues uninterrupted operation until the fault is resolved.

4.2. Experimental Validation

To fully verify the theoretical power flow characteristics, shared-half-bridge modulation principle, normal energy dispatching performance and energy-storage-coordinated HVRT mechanism analyzed in Section 2 and Section 3, a reduced-scale three-port single-stage PV–ESS experimental prototype is built in Figure 13. The prototype mainly consists of a battery energy storage (BES) unit, photovoltaic distribution box, grid-connected voltage regulator, grid-connected converter, dual-half-bridge (DHB) circuit, Buck/Boost unit and DSP controller. The battery energy storage unit simulates the energy storage port, and the photovoltaic distribution box provides input power for the photovoltaic port. The grid-connected voltage regulator is adopted to emulate various grid voltage disturbance conditions. The grid-connected converter together with the DHB forms the main three-port power conversion hardware, while the Buck/Boost unit assists energy regulation. The DSP controller is employed to implement system power management and HVRT control algorithms. All experimental parameters are consistent with the parameter setting rules adopted in the simulation platform, eliminating variable interference between simulation and physical test, which ensures the comparability of simulation and experimental results and fully supports cross-verification of theoretical derivation, simulation analysis and physical experiment. The unified key experimental parameters are defined as follows: steady-state DC Bus1 voltage u_Bus1 = 100 V, steady-state DC Bus2 voltage u_Bus2 = 50 V, auxiliary Port3 voltage adjustable range 0~160 V, steady PV output voltage u_PV = 100 V, high-frequency transformer turns ratio of DHB converter n = 2:1, switching frequency of all power switches set to 20 kHz.

Figure 14 presents the steady-state driving voltage and inductor current waveforms of the DHB converter under normal grid-connected operation mode, which fully matches the phase-shift modulation theoretical waveform given in Figure 5 of Section 2.1. The complementary driving signals of primary-side shared switches S_i3 and S_i4 and secondary-side switches S_i5 and S_i6 can be clearly observed from the oscilloscope curves, and an adjustable phase-shift angle exists between primary and secondary bridge arm drive pulses, which is consistent with the power transfer control mechanism derived in the theoretical part. By adjusting the phase difference φ of driving signals, bidirectional power flow between Bus1 and Bus2 is realized.

On the basis of steady-state modulation test, the soft-switching performance of the DHB converter is further tested, and the measured ZVS waveforms are shown in Figure 15. As analyzed in Section 2.1, the voltage across the power switch drops to zero before the turn-on pulse arrives, and zero-voltage turn-on is realized for all main power devices of DHB. The experimental phenomenon is completely consistent with the theoretical soft-switching condition deduction, which reduces switching loss and electromagnetic interference of the system.

Figure 16 records the dynamic energy dispatching response waveforms when PV grid-connected output power fluctuates, corresponding to the energy balance and power flow analysis in Section 2.2 and the normal operation simulation Case I in Section 4.1. When the PV output power exceeds the total power demand of grid and local loads, redundant power is transmitted from Bus1 to Bus2 through DHB converter to charge the energy storage unit; when PV generation is insufficient to support system load, energy storage discharges to supplement power deficit. The experimental waveform shows that the transition time of energy storage charge–discharge mode is less than 8 ms, with no obvious voltage overshoot or current oscillation on DC Bus1 and Bus2 during mode switching. The DC bus voltage fluctuation is controlled within ±3 V, which proves that the dual closed-loop control strategy proposed in Section 3.1 has an excellent dynamic suppression effect on PV intermittent power fluctuation.

As shown in Figure 17, when a high-voltage fault is detected on the grid side, Port3 begins to operate. At this point, the bidirectional Buck/Boost converter controls Port3 to raise its voltage, adjusting the voltage distribution on the DC bus so that PV output remains constant. The bus voltage u_Bus1 is elevated to a new voltage level to match the grid-side voltage under high-voltage fault conditions, enabling the system to continue operating continuously without disconnection until the fault is resolved.

It should be noted that this experimental platform adopts reduced-scale low-voltage parameters for principle-oriented verification. Although the core control logic and physical operating mechanism are completely retained, parasitic parameters including device junction capacitance and wiring inductance will differ from high-voltage industrial-grade equipment. These scaling-related influences need to be taken into consideration when conducting future practical engineering deployment.

In

Table 6. Quantitative experimental performance comparison of different schemes.

Scheme	Peak Efficiency	HVRT Response Latency	DC Bus Voltage Error
Conventional two-stage PV–ESS	94.1%	11 ms	±4.2 V
Single-stage HVRT from Ref. [3]	94.5%	12 ms	±3.7 V
Proposed three-port single-stage PV–ESS	95.2%	≤9 ms	±2.8 V

, the measured quantitative indices demonstrate that the proposed three-port single-stage PV–ESS achieves higher peak efficiency, faster HVRT dynamic response and smaller DC bus voltage deviation. It exhibits competitive overall performance compared with the two reference schemes.

Efficiency tests are performed for full-load PV-to-grid, PV-to-ESS charging, ESS-to-grid discharging, and HVRT transient conditions. The proposed scheme achieves 95.2%, 94.4% and 94.1% under the three steady-state modes, outperforming the conventional two-stage (94.1%, 93.3%, 93.0%) and traditional single-stage HVRT schemes (94.5%, 93.6%, 93.2%). During HVRT transients, deviation from the optimal soft-switching point reduces efficiency moderately to 93.5%, still higher than 92.2% (two-stage) and 92.7% (single-stage). Despite extra losses caused by circulating power in partial operating ranges, eliminating one DC–DC conversion stage brings dominant efficiency gains across most operating conditions.

Figure 18 shows the relationship between the maximum tolerable grid voltage swell and fault duration. For short-duration faults less than 0.4 s, the system can stably realize HVRT under 1.30 p.u. grid voltage swell. With the extension of fault duration, the energy storage port bears more cumulative transient power, so the maximum allowable overvoltage amplitude declines correspondingly. When the fault lasts 0.8 s, the maximum tolerable voltage swell drops to 1.18 p.u. The area below the boundary curve represents the reliable HVRT region.

5. Conclusions

This paper investigates a single-stage power management system for sustainable grid-connected photovoltaic power generation, develops a corresponding energy-storage-coordinated HVRT control strategy for grid voltage swell disturbances, and validates the rationality and fault-ride-through performance of the system and its control scheme through system-level simulation and reduced-scale laboratory prototype experiments. Different from conventional two-stage and multi-stage photovoltaic grid-connected structures, the proposed single-stage PV–storage grid-connected system adopts a three-port converter architecture that organically integrates photovoltaic power generation units, energy storage modules, and three-phase grid-connected converters, realizing flexible bidirectional power interaction among photovoltaic sources, energy storage equipment, and the utility grid. The designed system can implement accurate MPPT control for photovoltaic arrays by regulating the three-phase converter, which maximizes the utilization efficiency of solar energy resources and guarantees the sustainable power generation capability of the PV system. Meanwhile, the three-port converter allows the energy storage system to dynamically adjust its connection states and perform adaptive power scheduling in response to real-time grid operating conditions and photovoltaic output fluctuations, effectively suppressing power oscillations, smoothing intermittent PV output, and further improving the overall operational efficiency and grid-friendliness of the sustainable PV power generation system. For the common grid voltage swell fault in actual grid-tied scenarios, this work realizes an innovative energy-storage-coordinated HVRT operation mechanism. Under high-voltage fault conditions, the energy storage module participates in DC bus voltage modulation through the auxiliary port of the three-port converter, which actively elevates and stabilizes the DC bus voltage level. This technical solution effectively solves the problem of system off-grid operation caused by grid overvoltage faults, endowing the single-stage PV–storage grid-connected system with reliable high-voltage ride-through capability. Consequently, the continuous and stable grid-connected operation of the photovoltaic sustainable power generation system is fully guaranteed during fault duration. Both simulation analyses and physical experimental tests validate the rationality of the proposed single-stage system topology and the superiority of the energy-storage-coordinated HVRT control strategy. Compared with traditional improvement schemes that rely on additional hardware circuits or complex algorithm modification, the proposed method features a simpler system structure, lower hardware cost, and stronger operational practicability, which provides a feasible technical reference for the stable and sustainable grid-connected operation of photovoltaic power generation systems under complex grid fault conditions.

Nevertheless, this work still has several inherent limitations that deserve explicit acknowledgment. The ZVS soft-switching range is limited by phase-shift and duty-cycle modulation, failing to cover all operating conditions. The HVRT capability is dependent on ESS SOC margin, and transient fault operation imposes extra current stress on power devices and ESS. Bandwidth decoupling can eliminate routine control coupling between MPPT and HVRT loops, whereas slight dynamic interaction may still occur under extreme grid disturbances. In addition, the proposed strategy is mainly applicable to balanced voltage swell faults, with limited adaptability to unbalanced faults, composite grid disturbances and weak-grid scenarios. Furthermore, all validations are implemented on a low-voltage scaled-down prototype. The parasitic parameters and thermal characteristics of high-power industrial scenarios are not fully considered, which limits the direct scalability of the proposed scheme for practical grid-connected applications. Future work will focus on extending the ZVS range, improving fault adaptability, and conducting high-power experimental verification to bridge the gap between laboratory validation and field implementation.