Hierarchical Switching Control Strategy for Smart Power-Exchange Station in Honeycomb Distribution Network

Abstract

The Honeycomb Distribution Network is a new distribution network architecture that utilizes the Smart Power-Exchange Station (SPES) to enable power interconnection and mutual assistance among multiple microgrids/distribution units, thereby supporting high-proportion integration of distributed renewable energy and promoting a sustainable energy transition. To promote the continuous and reliable operation of the Honeycomb Distribution Network, this paper proposes a Hierarchical Switching Control Strategy to address the issues of DC bus voltage (U_dc) fluctuation in the SPES of the Honeycomb Distribution Network, as well as the state of charge (SOC) and charging/discharging power limitation of the energy storage module (ESM). The strategy consists of the system decision-making layer and the converter control layer. The system decision-making layer selects the main converter through the importance degree of each distribution unit and determines the control strategy of each converter through the operation state of the ESM’s SOC. The converter control layer restricts the ESM’s input/output active power—this ensures the ESM’s SOC and input/output active power stay within the power boundary. Additionally, it combines the Flexible Virtual Inertia Adaptive (FVIA) control method to suppress U_dc fluctuations and improve the response speed of the ESM converter’s input/output active power. A simulation model built in MATLAB/Simulink is used to verify the proposed control strategy, and the results demonstrate that the strategy can not only effectively reduce U_dc deviation and make the ESM’s input/output power reach the stable value faster, but also effectively avoid the ESM entering the unstable operation area.

1. Introduction

As a critical infrastructure supporting the global goals of low-carbon energy transition and carbon neutrality, the distribution network serves not only as a core platform for demand response management but also as a key carrier for accommodating massive distributed energy resources such as distributed generations (e.g., wind and solar power) and electric vehicles. Its operational efficiency and adaptability directly determine the progress of the energy system’s transition to a sustainable model, and it plays an irreplaceable role in alleviating reliance on fossil energy, addressing the energy crisis, and reducing carbon emissions [1,2]. The operation mode of traditional distribution networks is basically led by the supply side, with the feature of one-way radial power supply. Neither the planning and design stage nor the operational administration of traditional distribution networks is concerned with the integration of distributed energy resources. This model has significant shortcomings in terms of renewable energy sustainability [3,4]. With the constant increase in distributed generation penetration, the fast popularization of electric vehicles, and the stable growth of controllable loads, the current distribution network architecture can hardly satisfy users’ requirements in terms of environmental protection, dependability of electricity supply, electrical energy quality, and top-tier services [5,6].

In reference [7], to tackle the aforementioned problems, a novel AC–DC hybrid distribution network—dubbed the Honeycomb Distribution Network—was proposed, which is built on the power-electronic flexible interconnection device known as the Smart Power-Exchange Station (SPES). This network is capable of effectively balancing loads and improving power quality. Its hybrid DC–AC power distribution structure can adapt to the relevant DC loads and AC loads, respectively, in turn reducing energy conversion links, boosting efficiency, and alleviating harmonics. When a system fault occurs, It has the ability to swiftly isolate the fault and supply voltage and frequency support to the outage area. This network can improve energy utilization efficiency, establish a low-carbon energy system, help reduce carbon emissions in the energy sector, and support the achievement of sustainable development-related goals such as the "dual carbon" targets (carbon peaking and carbon neutrality). However, the stable operation of SPES is the key to realizing the advantages. In Reference [8], this study puts forward an adaptive control approach for islanded microgrid clusters with SPES as the basis. Despite the fact that this strategy can partially maintain the stable operation of SPES, it cannot effectively restrict the power and state of charge (SOC) of the energy storage module (ESM) in SPES within limits, and it is deficient in favorable dynamic response. In Reference [9], a hierarchical coordinated control strategy for SPES is proposed. This strategy can only manage SOC of the ESM to a certain extent; when the main converter power is inadequate, it loses the ability to manage SOC of the ESM, and it is only suitable for pure AC distribution systems. Hence, the control strategy for the stable operation of SPES still needs further study.

SPES’s flexible interconnect devices are mainly composed of the voltage source converters (VSC)/the soft open point (SOP) [10], the high-frequency isolation dual active bridge (DAB) [11], and the bidirectional buck–boost converter [12]. At present, numerous studies have been conducted to investigate the control strategies of these flexible interconnected devices. In reference [13], a flexible interconnection scheme with SOPs connected in parallel to tie switches is proposed, and a smooth switching control method is designed—this method can reduce voltage and current shocks to the system during load transfer. In addition, adding a certain percentage of the ESM’s devices to the system can further alleviate the unstable output of new energy and enhance the regulation capability of the distribution network [14]. In reference [15], researchers put forward a two-stage optimization method based on Interval Optimization, aiming to suppress voltage fluctuations and improve the economic benefits of distribution networks integrated with Energy Storage Integrated SOPs and high-penetration Distributed Generations. In reference [16], the study introduces a type of combined decentralized-local voltage control strategy targeting SOPs. This strategy enables rapid responses to frequent voltage fluctuations, and according to dynamic network partitioning results, it adjusts SOPs’ active power transfer between interconnected areas through decentralized optimization. In reference [17], a Moving Discretized Control Set–Model Predictive Control method for the DAB converter is proposed. This method is an ideal option for multi-objective control, with advantages including balancing two control targets effectively and exhibiting adaptive performance in terms of system impedance. In reference [18], a model predictive control-based method for optimizing the dynamic response of the bidirectional Buck–Boost converter’s DC bus voltage (U_dc) is proposed, which uses a large virtual capacitor to mitigate DC voltage changes when the load fluctuates abruptly. In reference [19], researchers investigate the control strategy of the grid-connected DC microgrid VSC under virtual capacitor control, which allows dynamic adjustment of the virtual capacitor size to provide inertial support for the DC power grid. The virtual capacitor algorithm of this strategy can adjust how the virtual capacitor responds to the voltage change rate and is suitable for various situations, but it does not have limiting functions. If the virtual capacitance value is too large, it might negatively impact the dynamic stability of the DC power grid. In reference [20], an improved adaptive virtual capacitor control is studied for the bidirectional Buck–Boost converter. When the load experiences disturbance, this control method is capable of adaptively regulating the size of the virtual capacitor according to the bus voltage change rate, thus increasing the DC bus inertia and reducing the speed of bus voltage change. While the virtual capacitor algorithm in this strategy exhibits limiting capability—a feature that helps circumvent the deficiencies noted in [19]—it lacks the ability to tune the virtual capacitor’s response speed to the U_dc change rate.

For SPES, it is essential to ensure that the input/output power and SOC of its ESM stay within limits during operation. Excessive charging and discharging of the ESM must be avoided, as such behavior would damage the ESM’s lifespan, reduce charging/discharging efficiency, and undermine system stability. To resolve this issue, we put forward a hierarchical switching control strategy with hard limiting capability. Based on the framework of the existing master-slave control strategy, this strategy incorporates the operating constraints of the ESM’s input/output power and SOC: it switches between control strategies according to the real-time operating states of the ESM’s input/output power and SOC, and restricts these parameters to keep the ESM operating stably within the allowed range. In addition, converters at each port of SPES require DC-side voltage support to function normally. As the shared DC side for all ports, the voltage stability of SPES’s DC bus is closely linked to the dependable operation of the entire system and is the key to maximizing SPES’s flexible regulation advantages. However, power electronic devices have inherent shortcomings of poor stability and low inertia. To suppress U_dc fluctuations in transient processes, this paper proposes a Flexible Virtual Inertia Adaptive (FVIA) control method. This method develops a new virtual capacitor control algorithm by referencing the ideas in [19,20]—this algorithm not only can adjust the response speed of the virtual capacitor but also has limiting capabilities. Moreover, to improve the power response speed, the FVIA control strategy in this paper not only includes current feedforward control but also designs a conduction coefficient. This coefficient ensures that virtual inertia support is maintained when the U_dc deviates from the rated value; during the voltage reverse recovery process, it keeps the virtual capacitance value extremely small, which in turn enhances the response speed of this process.

2. Honeycomb Distribution Network and SPES

2.1. Honeycomb Distribution Network Topology and Principle

Figure 1 shows the topology of a Honeycomb Distribution Network, which consists of multiple active microgrids or adjacent distribution units. At the power supply boundary of each unit, 6 SPES are arranged, and every SPES is linked to the common connection point (PCC) and power transmission line of the adjacent 3 units. The SPES serves as a hub for energy transmission, and each distribution unit can transmit energy with the connected SPES and can also mutual energy transfer with adjacent distribution units through the SPES [7].

Different operating scenarios of SPES:

(1)

If the active power of distributed energy sources, the ESM, and loads within a single distribution unit can realize internal balance, there will be no mutual active power transmission between the SPES and that unit;

(2)

When there is a power shortage or surplus within a distribution unit, it is necessary to absorb or supply active power to/from SPES and implement power dispatch through SPES;

(3)

If a unit experiences an internal fault, SPES can be used to realize load transfer—once the fault is cleared, power supply for loads beyond the fault range can be recovered.

2.2. Topology of SPES

Figure 2 shows the typical topology of SPES, which includes DC/AC or DC/DC bidirectional converters, the DC bus, the ESM, the control system, and communication equipment. The core function of this topology is to realize power mutual assistance between distribution units through the DC bus, and to control the converter’s operating state by means of the control system and communication equipment. Moreover, SPES has the ESM with a certain capacity, which can achieve flexible power flow adjustment and improve the operational stability of power distribution units. The DC bus of SPES is the foundation for SPES to decouple control strategies among nodes, and its voltage stability is of great significance to SPES’s stable operation. This stability is realized by the converters and their control circuits at each port, and the ESM port maintains the voltage stability in most scenarios.

As illustrated in Figure 3, the converters feature three distinct types of topologies: the DC/AC converter (VSC) [21], the DC/DC converter (DAB) [22,23], and the EMS converter (bidirectional Buck–Boost converter) [24,25].

3. Constraints for the SPES’s ESM

The capacity of the ESM does not need to be designed according to the capacity of the distribution unit but only needs to consider the possible power shortage between the distribution units [8,26]. The ESM in the SPES has the function of maintaining the U_dc, making up for the missing power, absorbing excess power, and adapting quickly to transient power disturbances.

For the ESM integrated into the SPES, its power during charging and discharging procedures can be expressed in the form:

$$P_{\mathrm{bat}}={\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}$$

(1)

where P_bat represents the charging/discharging power of the ESM and P_bat is positive value for discharge, while negative value for charging, and P_I denotes the power request value sent from unit I (where I = 1, 2, 3) to the SPES, and When P_I is positive, unit I is in power deficit and when P_I is negative, unit I is in power surplus.

Since the ESM maintains U_dc stability and enables flexible power dispatching, charging/discharging restrictions must be imposed based on its SOC, keeping power within a safe range. This ensures the ESM’s proper operation in the SPES while preventing overcharge or over-discharge. Meanwhile, the SOC of the ESM must be bounded by upper and lower thresholds. This ensures the ESM’s power output can satisfy the demands of sustaining the U_dc stability. The black solid line is the power run boundary as shown in Figure 4. It can be expressed as:

$$\left\{\begin{array}{l}{P}_{\min }\le P_{\mathrm{bat}}\le {\displaystyle \frac{U_{\mathrm{bat}}{Q}_{\mathrm{n}}(S-S_{\min })}{t}}, S_{\min }<S\le S_1\\ P_{\min }\le P_{\mathrm{bat}}\le P_{\max }, S_1<S\le S_2\\ {\displaystyle \frac{U_{\mathrm{bat}}{Q}_{\mathrm{n}}(S-S_{\max })}{t}}\le P_{\mathrm{bat}}\le P_{\max }, S_2<S\le S_{\max }\end{array}\right.$$

(2)

where S_max and S_min represent the SOC’s upper limit and lower limit for the ESM, respectively, S₁ and S₂ respectively function as the thresholds for restricting the charging rate and discharging rate of the ESM, P_max is the max discharging power of the ESM, -P_min represents the max charging power of the ESM, t denotes the duration of the output power, Q_n represents the capacity of the ESM, and U_bat represents the output voltage of the ESM.

When time t is fixed, the black slash ought to be a straight line—this line denotes the time from the operating point on the black slash until the ESM discharges to the lower SOC limit. The ESM must not continue outputting power once its SOC hits the lower limit, nor should it keep absorbing power when the SOC reaches the upper limit. Furthermore, when the SOC is about to reach the upper or lower limit, its charge/discharge rate should be constrained. In this paper, the range of the stable operation area is defined according to the SOC and P_bat of the ESM, as presented in Figure 4.

The light gray area in the figure is defined as the critical operation area, and this paper defines the range in which the center of the zone is 75% away from the boundary of the zone as the stable operation area (the dark gray area in the figure). When the SPES is in the stable operation area, it can realize stable power scheduling with each distribution unit. Therefore, there is no requirement to restrict the power of the ESM under this condition. In contrast, when the SPES operates in the critical operation area, it becomes essential to impose limits on the ESM’s charging and discharging power—this measure is intended to prevent the SPES from entering the unstable operation area.

4. SPES Hierarchical Switching Control Strategy Considering the ESM’s SOC

In the SPES, the ESM is critical to guaranteeing U_dc stability. When regulating input/output active power, corresponding control strategies are necessary to fulfill the ESM’s power and SOC operation boundary constraints, avoiding the ESM’s entry into the unstable operation zone. Transient power changes inside the SPES and the switching process of control strategies, however, will lead to U_dc fluctuations, reducing the SPES’s stability. Therefore, the SPES must be able to suppress U_dc fluctuations and respond swiftly to power changes. Against these problems, we adopt a hierarchical switching control strategy.

4.1. Overall Structure of SPES Hierarchical Switching Control Strategy

The SPES hierarchical switching control strategy is separable into the system decision-making layer and the converter control layer.

• The system decision-making layer determines the energy exchange relationship between distribution units by receiving their power deficits and surpluses, selects the main converter through the importance of each distribution unit, and chooses the weight coefficients of each converter through the operation status of the ESM’s SOC, to determine the control strategy of each converter;
• The converter control layer accepts the weight coefficients and energy exchange commands from the system decision-making layer and performs power scheduling through control measures. The voltage-stabilizing control converter automatically modifies the virtual inertia in line with the U_dc change rate (dU_dc/dt), thereby suppressing U_dc fluctuations and responding swiftly to changes in output power.

4.2. Control Strategy of the System Decision-Making Layer

The main converter is selected based on the importance of the road for each distribution unit. Since the main converter needs to adapt to the transient power fluctuations of the SPES when switching to constant DC voltage control, the non-faulty port with the lowest importance is chosen as the main converter, and the remaining ports serve as slave converters. If the main converter fails, re-selection is required using the same method. The weighting coefficients of each converter are shown in Figure 5, and the relationship between the weight coefficients of each converter and the control strategy is shown in Figures 9–11. To ensure the ESM’s power and SOC do not go beyond limits and avoid it entering an unstable operating region, this control strategy switches its control logic based on the ESM’s power and SOC operating states.

• When the ESM operates within the stable operation zone, it is capable of stable energy exchange with each distribution unit, and during this state, the ESM keeps the U_dc stable;
• When the ESM enters the critical operation zone, the ESM switches to constant power control to prevent the ESM’s power and SOC from exceeding limits and the ESM entering the unstable operation state. The power reference value meets $P_{\mathrm{batref}}={\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}$, and is restricted so that its output power is limited in the critical operating zone. Meanwhile, the main converter shifts from PQ control to the control mode of constant DC voltage to sustain the stable state of U_dc.

During the control strategy switching process, the weighting coefficients of each converter change at a specified rate from 0 to 1 or from 1 to 0. This method enables smooth switching between constant power control and constant DC voltage control, reducing the impact on the DC bus during the switching instant.

4.3. Control Strategy of the Converter Control Layer

4.3.1. FVIA Control Strategy

Given that the ESM converter is primarily responsible for maintaining U_dc stability in most SPES, this study uses the ESM converter as a case study to elaborate on the operational mechanism of FVIA control.

Introducing a virtual capacitor into the DC bus is an effective way to mitigate its voltage fluctuations [19,27]. What distinguishes FVIA control is its ability to flexibly and adaptively adjust the virtual capacitance value—this not only enhances the suppression of U_dc fluctuations but also accelerates the response speed of the ESM converter’s input/output power.

The current output of the ESM converter, along with its association with the virtual capacitor, is illustrated in Figure 6. When no virtual capacitor is introduced, the relationship between the dU_dc/dt and the current across the DC bus capacitor can be expressed as:

$$C_{\mathrm{dc}}{\displaystyle \frac{\mathrm{d}{U}_{\mathrm{dc}}}{\mathrm{d}t}}=i_{\mathrm{c}}=i_{\mathrm{out}}+\Delta i_{\mathrm{out}}-i_{\mathrm{in}}$$

(3)

where i_c refers to the capacitive current of the DC bus, i_out is the current output to the DC bus, ∆i_out represents the variation in output current when the DC bus is disturbed, and i_in is the output current of the ESM converter. When it runs at steady state, i_out = i_in, ∆i_out = 0 and i_c = 0. When it is disturbed, ∆i_out ≠ 0, and a virtual capacitor C_virb is introduced for suppressing the U_dc fluctuations. At this point, the relationship between dU_dc/dt and the current across the DC bus capacitor can be expressed as:

$$(C_{\mathrm{dc}}+C_{\mathrm{virb}}){\displaystyle \frac{\mathrm{d}{U}_{\mathrm{dc}}}{\mathrm{d}t}}=i_{\mathrm{c}}+\Delta i_{\mathrm{virb}}=i_{\mathrm{out}}+\Delta i_{\mathrm{out}}-i_{\mathrm{in}}$$

(4)

where ∆i_virb is the virtual capacitance current, C_virb is a virtual capacitor, and C_dc is a DC bus capacitor. It can be known from Equations (3) and (4) that with ∆i_out remaining unchanged, after introducing the virtual capacitor, (C_dc + C_virb) > C_dc, thereby reducing dU_dc/dt. From the above analysis, it can be determined that increasing the value of the virtual capacitor helps suppress U_dc fluctuations during the transient process and ensures the system operates stably.

For the purpose of further improving its dynamic operational performance, the FVIA control method is introduced in this paper, which couples dU_dc/dt with C_virb, so that when disturbance occurs, the size of C_virb can be adaptively adjusted according to the size of dU_dc/dt to change its own inertia. To meet the requirements of various voltage stability and avoid the loss of stability of the system due to excessive virtual inertia, a function expression with adjustable response speed and limiting ability is proposed, as shown in (6).

The expression for the virtual capacitor current ∆i_virb is:

$$\Delta i_{\mathrm{virb}}=C_{\mathrm{virb}}{\displaystyle \frac{\mathrm{d}{U}_{\mathrm{dc}}}{\mathrm{d}t}}=-C_{\mathrm{virb}}{\displaystyle \frac{\mathrm{d}(U_{\mathrm{dcref}}-U_{\mathrm{dc}})}{\mathrm{d}t}}=-{\displaystyle \frac{C_{\mathrm{virb}}}{T_2}}\left[(U_{\mathrm{dcref}}-U_{\mathrm{dc}})-(U_{\mathrm{dcref}}-U_{\mathrm{dc}0})\right]$$

(5)

where U_dcref is the U_dc reference value (generally the rated value), U_dc0 is the U_dc of the last sample, and T₂ is the sampling period.

The expression for the virtual capacitor C_virb is:

$$C_{\mathrm{virb}}=\left\{\begin{array}{l}{C}_{\mathrm{v}0},\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em} N_1\cdot N_2=0\\ C_{\mathrm{v}0}+m(1-e^{-a\left|{\displaystyle \frac{\mathrm{d}{U}_{\mathrm{dc}}}{\mathrm{d}t}}\right|}), N_1\cdot N_2\ne 0\end{array}\right.$$

(6)

where m represents the adjustment coefficient for the virtual capacitance, a is the response coefficient, C_v0 denotes the virtual capacitance value under steady-state conditions, and N₁ and N₂ are the conduction coefficients.

To ensure system stability during steady-state operation, a threshold value K is established for the dU_dc/dt. This threshold prevents the frequent switching of C_virb that could otherwise be triggered by minor U_dc oscillations. The determination of K is based on the maximum dU_dc/dt value observed within the maximum anticipated oscillation range of the U_dc under steady-state conditions [20].

In the selection of the response factor a, when the capacity of each connected distribution unit is large and the disturbances are small, a larger inertia support is required to better reduce the U_dc fluctuations. In this instance, it is necessary to raise the value of a to raise the virtual capacitance value when dU_dc/dt is small. The trend of C_virb with a is shown in Figure 7a. Conversely, the virtual capacitance value can be appropriately reduced.

When ∆₁ = U_dc − U_dcref and dU_dc/dt are of the same sign, The U_dc is deviating from its rated value, and at this time, N₂ is in the conduction state to suppress the voltage offset. when ∆₁ = U_dc − U_dcref and dU_dc/dt are of the different sign, the U_dc is in the reverse recovery stage, and at this time N₂ is in the cut off state to restore the original response speed. Since the U_dc boosts the system’s virtual inertia when deviating from the rated value, the voltage offset is lessened. As a result, the voltage recovery amount in the reverse recovery stage is reduced, which shortens the recovery process and improves the recovery speed. The dynamic performance can be improved by appropriately selecting the value of m while quickly responding to changes in output power. The variation trend of C_virb with m is shown in Figure 7b.

The U_dc is processed by an equivalent Washout filter to obtain dU_dc/dt. T₁ represents the time constant of the Washout filter [20]. The control principle is shown in Figure 8 and Figure 9.

4.3.2. Control Strategies for Each Converter of ESS

First, the control strategy of the ESM converter is presented. Its constant DC voltage control incorporates current feed-forward control and FVIA control measures on the basis of voltage-current double closed-loop, as illustrated in Figure 9.

The output current of the half-bridge ESM converter is denoted by i_bat, and the duty cycle is designated as D. The sum of currents absorbed by each converter from the DC bus is filtered by a low-pass filter to obtain i_mgsum. After adding i^*_batref and ∆i_virb, the result is multiplied by the weighting coefficient (1-y) to obtain the DC bus current reference value i_batref1. The inclusion of current feedforward control can further enhance the response speed of voltage and power [8].

The expression for the power-limiting module in the constant power control state is:

$$P_{\mathrm{batref}}=\left\{\begin{array}{l}{P}_{\min }\le {\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}\le {\displaystyle \frac{Q_{\mathrm{n}}(S-S_{\min })}{t}}, S_{\min }<S\le S_1\\ P_{\min }\le {\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}\le P_{\max }, S_1<S\le S_2\\ {\displaystyle \frac{Q_{\mathrm{n}}(S-S_{\max })}{t}}\le {\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}\le P_{\max }, S_2<S\le S_{\max }\end{array}\right.$$

(7)

where U_bat is the output voltage of the ESM, and i_bat is the output current of the ESM. After inputting ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}$, power limitation is performed through Equation (7) to output the power reference value P_batref. When ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}$ is within the power boundary, $P_{\mathrm{batref}}={\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}$. When ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}$ exceeds the power boundary, P_batref equals the power boundary value. The output P_batref is divided by U_bat and then multiplied by the weighting coefficient y to obtain the current reference value i_batref2. The final current reference value i_batref is obtained by i_batref1 + i_batref2. Through this power-limiting control measure, the system is operated within the power boundary to prevent the ESM from entering the unstable operation area.

Then, the following is an explanation of the control strategies for the VSC. U_dcQ control is applied in the constant DC voltage control, and FVIA control is additionally included in the control process. The constant power control uses PQ control, enabling bidirectional power transmission, as shown in Figure 10. The output V_d^* and V_q^* will control the conduction and shutdown of the six bridge arms through the modulation strategy. As the SPES performs only active power interaction, the reactive power reference value is 0 (i_qref = 0).

Finally, the control scheme applied to the DAB converter is detailed as follows. The DAB employs a single-phase-shift modulation strategy, which offers advantages such as fewer control parameters and favorable stability [28]. Its constant DC voltage control adopts voltage-current double closed-loop control, ensuring the U_dc remains stable during bidirectional power flow. FVIA control is also added, as shown on Figure 11. The constant power control adopts open-loop control due to the low requirement of power control accuracy. The calculation formula for its phase-shift ratio is:

$$d_2^{\prime }=\left\{\begin{array}{l}{\displaystyle \frac{1}{2}}-\sqrt{{\displaystyle \frac{1}{4}}-{\displaystyle \frac{2L{f}_{\mathrm{s}}{P}_{\mathrm{ref}}}{V_1{V}_2}}}, {\displaystyle \frac{N{V}_1{V}_2}{8L{f}_{\mathrm{s}}}}\ge P_{\mathrm{ref}}>0\\ -{\displaystyle \frac{1}{2}}+\sqrt{{\displaystyle \frac{1}{4}}+{\displaystyle \frac{2L{f}_{\mathrm{s}}{P}_{\mathrm{ref}}}{V_1{V}_2}}}, -{\displaystyle \frac{N{V}_1{V}_2}{8L{f}_{\mathrm{s}}}}\le P_{\mathrm{ref}}\le 0\end{array}\right.$$

(8)

where the meanings of these symbols (related to the DAB converter) are explained in

Table A1. The symbols and their corresponding meanings.

Symbol	Meaning
T1, T2	Inertial time constant
img1	Port A outputs current to the DC bus
img2	Port B outputs current to the DC bus
img3	Port C outputs current to the DC bus
P	The actual value of the power of the VSC
V1	The primary side voltage of the DAB
V2	The secondary side voltage of the DAB
L	DAB inductor
fs	DAB frequency
N	DAB’s transformer ratio
i1	The current input on the primary side of the DAB

of the Appendix A.

5. Simulation Verification

For validating the effectiveness of the hierarchical switching control strategy tailored for SPES put forward in this study, a simulation model featuring a SPES connected to three distribution units was constructed using MATLAB/Simulink. Among them, Port A is a DC distribution unit, and Ports B and C are AC distribution units. As shown in Figure 12 is the simulation model, and the parameters applied in the simulation are outlined in

Table 1. Parameters of the simulation system.

Parameter	Numerical Value
Rated voltage of DC distribution unit UdN	750 V
AC distribution unit rated voltage UaN	380 V
Rated voltage of the ESM UbatN	400 V
DC bus rated voltage UdcN	750 V
Rated frequency of AC distribution units fN	50 Hz
Rated capacity of DC distribution unit PdN	50 kW
Rated capacity of AC power distribution unit PaN	130 kW
Rated capacity of the ESM Qn	250 A∙h
DC bus capacitance Cdc	20 mF
The ESM converter inductor Lbat	3 mH
Virtual capacitor steady state value Cv0	2 mF

.

A simulation test under the condition of SOC = 50% was designed to verify the effectiveness of the system decision-making layer control and the FVIA control strategy, respectively. Only current feedforward control is included in the control strategy of [8]. By contrast, the FVIA control method, after adding current feedforward control, further incorporates the improved FVIA control based on [19,20]. Hence, this study contrasts the proposed control strategy with the one presented in [8] as well as the adaptive virtual capacitor control strategy in [20].

5.1. Simulation Verification of the System Decision-Making Layer Control Strategy

5.1.1. Verification of Switching Control Strategy for DAB as the Main Converter

For verifying the DAB control strategy’s effectiveness, Port A is set as the main converter, with the other two ports (Ports B and C) configured as slave converters. From t = 0.1 s to 0.3 s, Port A has a power surplus of 10 kW, while Ports B and C both have a power deficit of 30 kW. At this point ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}=50 \mathrm{kW}}$. The ESM operates in the stable region (y = 0), and the U_dc stability is maintained by the ESM converter. At t = 0.3 s, the load of Port B suddenly increases by 30 kW, as shown in Figure 13b. At this point ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}=80 \mathrm{kW}}$. When the ESM enters the critical operation area (y = 1), the DAB switches to voltage-stabilizing control to sustain U_dc stability, while the ESM converter shifts to constant power control. As depicted in Figure 13a, after the control strategy switches at 0.3 s and undergoes a brief transient process, the U_dc rapidly stabilizes within a reasonable range and operates steadily in the critical region.

5.1.2. Verification of Switching Control Strategy for VSC as Main Converter

For verifying that the VSC control strategy is effective, Port C is set as the main converter, with the other two ports configured to act as slave converters. From t = 0.1 to 0.3 s, Ports A and C have power deficits of 10 kW and 30 kW, respectively, while Port B has a 10 kW power surplus. At this point ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}=30 \mathrm{kW}}$. The ESM operates in the stable region (y = 0), and the stability of the U_dc is maintained by the ESM converter. At t = 0.3 s, the load of Port B suddenly increases by 70 kW, as shown in Figure 14b, and the ESM reaches the power boundary (y = 1). At this point ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}=100 \mathrm{kW}}$. Meanwhile, the VSC shifts to voltage-stabilizing control to keep the U_dc stable, and the ESM converter switches to constant power control. As shown in Figure 14a, after a brief transient process and the switching of the control strategy, the U_dc stabilizes at the power boundary. This indicates that the proposed control strategy, with the VSC as the main converter, can similarly maintain U_dc stability when the ESM operates in the critical region.

5.2. Simulation Verification of FVIA Control Strategy

To validate the FVIA control strategy’s effectiveness, an experimental setup was built, using the ESM converter as the research example. In this experiment, the ESM runs in the stable region (y = 0)—a situation where maintaining the U_dc stability is the responsibility of the ESM converter. During the period from t = 0.1 to 0.2 s, Ports A and B each have a 30 kW power surplus, and Port C has a 30 kW power deficit. At this point ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}=-30 \mathrm{kW}}$. At t = 0.2 s, port B load suddenly increases by 90 kW. At this point ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}=60 \mathrm{kW}}$. As shown in Figure 15a are the voltage fluctuations of the three control strategies. The yellow curve represents the control strategy from [8], the cyan curve represents the control strategy from [20], and the magenta curve represents the control strategy proposed in this paper.

As the figure clearly reveals, at every instant of load variation, the U_dc fluctuates with the smallest amplitude under the FVIA control strategy put forward in this paper. Take the sudden rise in the Port B load at 0.2 s as an example: the voltage fluctuation range of both the proposed strategy and the control strategy in [20] is noticeably narrower than that of the control strategy in [8]. Moreover, the proposed strategy could recover stability faster than the strategy in [20]. Simultaneously, as shown in Figure 15b, under the proposed control strategy, the output power of the ESM port could reach stability more quickly. This fully illustrates that the strategy can effectively restrain U_dc fluctuations and elevate the system’s voltage and active power dynamic response speed during load variations.

Based on the disturbance scenario where the load at Port B surges by 90 kW at 0.2 s, the voltage and power response indicators of the three strategies were calculated, and

Table 2. Accuracy Indicators of Each Control Strategy.

Control Strategy	Voltage ITAE (V·s)	Power ITAE (kW·s)	Voltage Recovery Time (s)	Power Stabilization Time (s)
Proposed FVIA Control in This Paper	0.82	3.68	0.03	0.032
Literature [20]	1.35	5.21	0.05	0.062
Literature [8]	2.17	7.83	0.06	0.071

shows the results. It is clear that each performance indicator of the control strategy proposed herein is better than that of the control strategies in [8,20].

5.3. Simulation Validation of Power-Limiting Measures for the ESM

For confirming the effectiveness of the ESM’s power limitation measures, port A has a power surplus of 10 kW from t = 0.1 to 0.3 s, while Ports B and C have power deficits of 40 kW and 30 kW, respectively. At this point ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}=80 \mathrm{kW}}$ (y = 1). Port C is set as the main converter. At t = 0.3 s, the load of Port B suddenly increases by 30 kW. At this point ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}=110 \mathrm{kW}}$.

According to Equation (7), at this moment P_batref =100 kW. Figure 16 shows ${\displaystyle \sum _{\mathrm{I}=1,2,3}{P}_{\mathrm{I}}}$, P_batref and the actual active power output of the ESM, and the output power of the ESM stabilizes at the power boundary. The output power of the ESM stabilizes at the power boundary. This control measure effectively prevents the ESM from entering the unstable region.

6. Conclusions

Addressing the issues of U_dc fluctuations, the ESM’s SOC, and charging/discharging power over-limit faced by the SPES in the Honeycomb Distribution Network during operation, this paper proposes a hierarchical switching control strategy. By virtue of the collaborative operation between the system decision-making layer and the converter control layer, coupled with the FVIA control method, this strategy effectively enhances the stability, dynamic performance, and power response speed of the SPES. The main conclusions are as follows:

(1) The hierarchical switching control strategy proposed in this paper successfully avoids over-limits of the ESM’s charging/discharging power and SOC. It effectively prevents the ESM system from entering the unstable operation zone and extends the ESM’s service life.

(2) By coupling with the U_dc change rate, the proposed FVIA control method adaptively adjusts the virtual capacitor size, a measure that significantly enhances system inertia. This enables the U_dc to remain stable even amid load disturbances, thereby effectively improving the system’s dynamic response to voltage deviations.

(3) Leveraging FVIA’s characteristic—which increases virtual inertia to suppress deviation whenever the voltage departs from its rated value and reduces it to accelerate recovery once restored—the input/output power can quickly be stabilized. This addresses the trade-off between voltage stability and power response speed in traditional control methods.

(4) The model established in this paper is built on a single SPES connected to three distribution units. It does not take into account more complex scenarios of honeycomb distribution networks with multiple interconnected SPES, such as multi-SPES coordinated control, power distribution optimization, and global stability analysis.

The hierarchical switching control strategy proposed herein serves as an effective solution to ensure the stable and highly efficient operation of the Honeycomb Distribution Network. Tackling its current limitations and conducting in-depth research in the indicated future directions will substantially promote the maturation and application of technology related to the Honeycomb Distribution Network. It should be noted that future research needs to further explore the coordinated operation mechanism among multiple SPES in large-scale honeycomb networks, so as to enhance the resilience, economy, and new energy accommodation capacity of the entire network.