Control method of subdivision type energy storage system

Abstract

The invention discloses a subdivision type energy storage system control method. The method comprises the following steps: acquiring operation state parameters and external control requirements of a plurality of energy storage subdivisions in real time; generating a control signal for controlling the on-off state of each switch unit array according to a topology control algorithm based on the operation state parameters and external control requirements; and the on-off state of the intelligent power electronic switch matrix is dynamically adjusted according to the control signal, and the electrical connection relation among the energy storage sub-cabins is reconstructed. By deploying an intelligent power electronic switch matrix and executing a topology control algorithm, millisecond-level dynamic reconstruction of an electrical connection relationship among a plurality of energy storage sub-cabins is realized, and the technical problems that due to fixed electrical connection, system operation flexibility is insufficient, faults cannot be thoroughly isolated, and virtual energy storage units are difficult to construct as required are solved.

Description

Technical Field

[0001] This invention relates to the field of energy storage device control technology, and in particular to a control method for a compartmentalized energy storage system. Background Technology

[0002] With the increasing penetration of renewable energy and the growing demand for peak shaving and frequency regulation on the grid side, electrochemical energy storage systems are widely used due to their flexible power regulation capabilities. To improve system capacity, achieve modular design, and facilitate operation and maintenance, modular energy storage systems have become the mainstream technology. These systems typically consist of multiple independent energy storage compartments (or energy storage units), which are physically isolated but electrically integrated to the same DC bus or AC grid through specific connection methods.

[0003] In existing technologies, the electrical connection topology of compartmentalized energy storage systems is mostly a fixed structure. A common practice is to directly connect all energy storage compartments in parallel to a common DC bus, or to combine a fixed number of compartments in series before paralleling them. While this pre-defined fixed connection method is simple and reliable in system design, it reveals significant limitations in actual operation. First, key parameters such as the system's output voltage and power rating are fixed by the physical connection, making dynamic adaptation to real-time grid dispatch instructions (such as high power response for frequency regulation and high energy capacity for peak shaving and valley filling) impossible. This results in insufficient operational flexibility and makes it difficult to maximize revenue across various services in the electricity market.

[0004] A more prominent problem is that the fixed topology severely restricts the system's reliability and lifespan. When a storage compartment in the system experiences performance degradation or fails due to internal cell differences, aging, or external factors, severe circulating currents are generated in a parallel structure. This faulty compartment becomes the "weak link" of the entire system, dragging down the overall output capacity and potentially triggering a chain reaction. In a series structure, it can paralyze the entire branch, causing a sharp drop in system capacity. Existing solutions mostly focus on passively disconnecting the faulty compartment from the system after a fault occurs using circuit breakers or contactors. However, this is a "post-event remedy," and the disconnection action itself may cause grid shocks, without fundamentally changing the inherent defects of the system architecture.

[0005] Furthermore, existing control methods for compartmentalized energy storage systems are mostly focused on balancing strategies at the battery management unit level and power allocation algorithms at the energy management system level. Their core is optimization based on unchanging physical connections. This "soft" control cannot break free from the constraints of "hard" connections and cannot achieve real-time reconstruction of the compartmentalized physical topology. Therefore, it is difficult to simultaneously achieve absolute isolation under fault conditions, flexible configuration under different application scenarios, and optimization of efficiency throughout the system's entire lifecycle. Summary of the Invention

[0006] The purpose of this invention is to provide a control method for a compartmentalized energy storage system. By deploying an intelligent power electronic switch matrix and executing a topology control algorithm, the method achieves millisecond-level dynamic reconfiguration of the electrical connection relationships between several energy storage compartments. This solves the technical problems of insufficient system operation flexibility, inability to completely isolate faults, and difficulty in constructing virtual energy storage units on demand caused by fixed electrical connections.

[0007] To address the aforementioned technical problems, a first aspect of this invention provides a control method for a compartmentalized energy storage system. The compartmentalized energy storage system includes a central controller, several energy storage compartments, and a smart power electronic switch matrix. The smart power electronic switch matrix is ​​connected between the several energy storage compartments and includes several switch unit arrays. The central controller controls the on / off state of each switch unit in the smart power electronic switch matrix. The control method is applied to the central controller and includes the following steps: Real-time acquisition of the operating status parameters and external control requirements of the aforementioned energy storage compartments; Based on the operating status parameters and the external control requirements, a control signal for controlling the on / off state of each of the switching units is generated according to the topology control algorithm. Based on the control signal, the on / off state of the intelligent power electronic switch matrix is ​​dynamically adjusted to reconstruct the electrical connection relationship between the several energy storage compartments.

[0008] Furthermore, the state parameters include at least one of state of charge, health state, temperature, and internal resistance, and the external control requirements include grid dispatch instructions, power allocation requirements, or fault handling instructions. Based on the operating state parameters and the external control requirements, and according to the topology control algorithm, control signals are generated for controlling the on / off state of the intelligent power electronic switch matrix, including: Analyze the external control requirements to determine the unit performance requirements of one or more of the virtual energy storage units to be formed, wherein the unit performance requirements include at least one of target output voltage, target output power and operating mode; Based on the operating status parameters, available compartments are selected from the plurality of energy storage compartments to form a set of available compartments. The selection criteria include that the state of charge is in the effective working range, the health status is higher than a preset threshold, and there are no fault signs. With the goal of matching the unit performance requirements and maximizing the overall system efficiency, the energy storage compartments in the available compartment set are combined and optimized to obtain the optimal compartment combination scheme that satisfies the target electrical connection topology. Based on the optimal compartment combination scheme, a corresponding switching timing logic table is generated.

[0009] Furthermore, the step of combining and optimizing the energy storage compartments in the available compartment set to calculate the optimal compartment combination scheme that satisfies the target electrical connection topology includes: Based on the unit performance requirements and the available compartment set, a candidate topology configuration set is generated, wherein each candidate topology configuration defines the series and parallel connection relationship of the energy storage compartments; For each candidate topology configuration in the candidate topology configuration set, the performance consistency index of any virtual energy storage unit within the candidate topology configuration is calculated. The performance consistency index is calculated based on the variance of the state of charge and health status of all energy storage compartments constituting the virtual energy storage unit. Calculate the system redundancy of the candidate topology configuration, wherein the system redundancy is quantified in the following way: after simulating the failure and isolation of any energy storage compartment, the maximum number of independent combinations of virtual energy storage units that meet the unit performance requirements can be formed by topology reconstruction in the remaining available energy storage compartments; The candidate topology configuration that simultaneously satisfies the performance consistency index being higher than a first preset threshold and the system redundancy index being higher than a second preset threshold is selected as the optimal compartment combination scheme.

[0010] Further, calculating the performance consistency index of any virtual energy storage unit within the candidate topology configuration includes: Obtain the real-time internal resistance and temperature parameters of all energy storage compartments that constitute the virtual energy storage unit; Based on the state of charge, health status, real-time internal resistance and temperature parameters, calculate the maximum sustainable discharge power or maximum sustainable charging power of each energy storage compartment in the current operating mode. The coefficient of variation of the maximum sustainable discharge power or the maximum sustainable charging power of all energy storage compartments in the virtual energy storage unit is used as a dynamic power consistency index, wherein the coefficient of variation is the ratio of the standard deviation to the mean. The static consistency index, calculated based on the variance of state of charge and health status, is weighted and fused with the dynamic power consistency index to generate a comprehensive performance consistency index, wherein the weighting coefficients are dynamically adjusted according to the current operating mode of the virtual energy storage unit.

[0011] Furthermore, the step of combining and optimizing the energy storage compartments in the available compartment set to calculate the optimal compartment combination scheme that satisfies the target electrical connection topology includes: Obtain the current electrical connection topology of the compartmentalized energy storage system as the initial topology; Based on the unit performance requirements and the available compartment set, a set of all feasible target topologies that can be reached from the initial topology through a single reconfiguration operation is generated, wherein a single reconfiguration operation refers to a single topology switch achieved by changing the on / off state of the intelligent power electronic switch matrix. For each target topology in the set of feasible target topologies, calculate the reconstruction cost required to get from it to the initial topology, where the reconstruction cost is a weighted sum of the number of switching units that need to change state and the corresponding switching action time constant. From the set of feasible target topologies, the target topology that can meet the unit performance requirements and has the lowest cost of the reconfiguration action is selected as the optimal compartment combination scheme.

[0012] Further, generating a set of all feasible target topologies reachable from the initial topology through a single reconstruction operation includes: The initial topology is subjected to a switch operation safety prediction, and all switch state combinations that would cause a short circuit or overcurrent in the power device are identified and eliminated to obtain a safe switch operation space. Within the safety switch action space, all single switch state changes are enumerated to generate a first-layer transition topology state set. For each transition topology in the first layer of transition topology state set, electrical parameters are verified based on the unit performance requirements. Transition topologies whose output voltage deviation from the target output voltage is within the allowable range and whose loop current does not exceed the preset safety threshold are selected to obtain the second layer of candidate topology set. Based on the second-layer candidate topology set, the single switch state change is enumerated again, and the safety prediction and electrical parameter verification are repeated until all generated topology states meet the unit performance requirements. The union of these topology states is taken as the feasible target topology set.

[0013] Furthermore, the control signal includes a fault isolation control signal; The step of dynamically adjusting the on / off state of the intelligent power electronic switch matrix according to the control signal and reconstructing the electrical connection relationship between the plurality of energy storage compartments includes: The unique identifier of the faulty energy storage compartment to be isolated is parsed from the fault isolation control signal; Query the system topology connection mapping table to determine the set of target switch units corresponding to all electrical connection paths with the faulty energy storage compartment; A fault isolation execution instruction sequence is generated, which controls all switch units in the target switch unit set to synchronously switch to the open state, thereby achieving multi-path electrical isolation of the faulty energy storage compartment; After the preset isolation verification time window, the status register of the target switch unit set and the insulation impedance parameters of the system's total circuit are read back to verify that the faulty energy storage compartment has been completely isolated and has no residual leakage current.

[0014] Furthermore, the control signal includes a virtual unit control signal; The step of dynamically adjusting the on / off state of the intelligent power electronic switch matrix according to the control signal and reconstructing the electrical connection relationship between the plurality of energy storage compartments includes: The target topology connection diagram contained in the virtual unit control signal is analyzed. The target topology connection diagram defines the connection relationship between several target energy storage compartments in the form of a node adjacency matrix. The target topology connection diagram is compared with the current topology state of the system to generate a differentiated set of switching action instructions, which includes the switching units whose states need to be changed and their target states. The switching action instruction set is executed in stages according to the order of first establishing the series branch and then establishing the parallel connection, and a waiting interval of no less than the power device recovery time is introduced after each stage operation is completed. After all switching actions have been completed, the consistency between the actual electrical connection relationship and the target topology connection diagram is verified by sampling the connection point voltage and loop current direction of each target energy storage compartment.

[0015] Furthermore, after generating the corresponding switching timing logic table based on the optimal compartment combination scheme, the process further includes: An independent operation control loop is configured for the at least one virtual energy storage unit, wherein each operation control loop calculates and executes the charging and discharging power command of the virtual energy storage unit based on the unit performance requirements of its corresponding virtual energy storage unit. The voltage and frequency fluctuations of the bus connecting each virtual energy storage unit are monitored in real time. When the fluctuations exceed the preset stable range, the damping power command is dynamically allocated based on the remaining regulation capacity of at least one of the virtual energy storage units. The damping power command is superimposed with the original charging and discharging power command to generate the final power command of at least one of the virtual energy storage units, and executed through the operation control loop to achieve coordinated and stable control of the system bus by at least one of the virtual energy storage units.

[0016] Furthermore, after acquiring the operating status parameters of the plurality of energy storage compartments in real time, and before generating a control signal for controlling the on / off state of each of the switching units based on the operating status parameters and the external control requirements according to the topology control algorithm, the method further includes: Based on the historical temperature data and real-time internal resistance data in the operating status parameters, the internal resistance growth trend and capacity decay trajectory of each energy storage compartment in the future preset period are predicted by the electrochemical aging model. Identify energy storage compartments whose internal resistance growth trend slope or capacity decay rate exceeds the adaptive early warning threshold and mark them as potential risk compartments. When the topology control algorithm is executed subsequently, an operational constraint strategy is applied to the potential risk compartment. The operational constraint strategy includes limiting its maximum charge and discharge current, avoiding placing it in a virtual energy storage unit with high power demand, or reducing its priority in combinatorial optimization.

[0017] Accordingly, a second aspect of the present invention provides an electronic device, including: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform the above-described compartmentalized energy storage system control method.

[0018] Accordingly, a third aspect of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-described compartmentalized energy storage system control method.

[0019] The above-described technical solutions of the embodiments of the present invention have the following beneficial technical effects: 1. By deploying an intelligent power electronic switch matrix and executing topology control algorithms, the system completely breaks free from the constraints of fixed electrical connections, enabling dynamic reconfiguration of the connection relationships of each energy storage compartment within milliseconds. This not only allows the system to "fission" into multiple virtual energy storage units with different voltages and power ratings according to grid demand, simultaneously undertaking different application tasks, but also achieves complete physical isolation of faulty compartments, fundamentally eliminating fault propagation paths and solving the key problems of rigid operating modes and reliability limitations imposed by fixed topologies in existing systems; 2. Beyond traditional simple power allocation, the optimization process simultaneously considers static performance balance, dynamic power capability, system redundancy, and the cost and safety of the reconfiguration process itself. This multi-objective and multi-constraint optimization ensures that the topology selected by the system under any operating condition is not only the most efficient at present, but also the most resilient to future failures, and the safest and most economical for the switching process. This significantly extends the service life of the equipment while improving the overall efficiency of the system. 3. It not only includes accurate and rapid isolation and verification processes after a fault and closed-loop safety verification during the construction of virtual units, but also innovatively introduces predictive maintenance strategies based on electrochemical models and control mechanisms for the coordinated stabilization of the power grid by multiple virtual units. This forms a complete technical closed loop from cell-level potential risk warning to network-level reconfiguration safety execution, and then to system-level proactive support, elevating system safety management from "passive response to faults" to a new level of intelligence of "proactive prediction, proactive defense, and proactive support". Attached Figure Description

[0020] Figure 1 This is a flowchart of the control method for a compartmentalized energy storage system provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0022] Please refer to Figure 1 The first aspect of the present invention provides a control method for a compartmentalized energy storage system. The compartmentalized energy storage system includes a central controller, several energy storage compartments, and an intelligent power electronic switch matrix. The intelligent power electronic switch matrix is ​​connected between the several energy storage compartments and includes several switch units. The central controller controls the on / off state of each switch unit in the intelligent power electronic switch matrix.

[0023] Specifically, the compartmentalized energy storage system adopts a three-layer hardware architecture design, forming a complete control and execution system. The core components of the system include a central controller, energy storage compartment clusters, and an intelligent power electronic switch matrix, which work together organically through a hierarchical control architecture. The central controller, acting as the system's brain, employs a multi-core processor architecture, runs a real-time operating system, and connects to various underlying devices via CAN bus, Ethernet, or fiber optic communication interfaces. The controller not only executes complex topology control algorithms but also possesses comprehensive functions such as data acquisition, status monitoring, fault diagnosis, and communication coordination. Its internal storage includes a system topology connection mapping table, a switch unit status table, and an energy storage compartment parameter database, providing data support for control decisions. The energy storage compartments are the system's energy storage units. Each compartment adopts a standardized, modular design, containing a battery pack, a battery management system, a bidirectional converter, and a local controller. The battery packs typically use lithium-ion batteries or flow batteries, and their DC output terminals are connected to the corresponding ports of the intelligent power electronic switch matrix via power cables. The battery management system monitors parameters such as voltage, current, and temperature within each energy storage compartment in real time, calculates the state of charge and health status, and uploads the status data to the central controller via a communication interface. The intelligent power electronic switch matrix is ​​the key hardware device for achieving dynamic topology reconfiguration in this invention. This matrix consists of several switch units arranged in an n×m array. Each switch unit includes a fully controlled power electronic switch device (such as an IGBT or MOSFET), a drive circuit, a status detection circuit, and a communication interface. The input ports of the switch matrix are connected to the output ports of each energy storage compartment, and the output ports are connected to the system DC bus or load. By controlling the on / off states of different switch units, different electrical connection paths can be established between energy storage compartments, realizing various topologies such as series, parallel, and mixed connections.

[0024] Accordingly, the control method applied to the central controller includes the following steps: Step S100: Real-time acquisition of operating status parameters and external control requirements of several energy storage sub-compartments.

[0025] During the operation of the modular energy storage system, the central controller continuously collects dynamic operating status parameters of each energy storage module through the underlying sensor network and communication bus. These parameters include, but are not limited to, the state of charge reflecting the remaining power, the health status characterizing the degree of battery aging, real-time temperature monitoring data, and DC internal resistance changes. Simultaneously, the central controller receives external control requests from the upper-level energy management system or grid dispatch center through a standard communication interface. These requests may include grid frequency regulation commands requiring a response, power allocation requirements for specific time periods, or fault handling commands triggered internally by the system. This step establishes the system perception layer, providing a complete and real-time data foundation for subsequent intelligent decision-making, ensuring that control strategies accurately reflect changes in the system's internal state and external demands. In practical applications, when the system participates in grid frequency regulation services, this step frequently collects power output capacity data from each module; when the system is performing peak shaving and valley filling, it focuses on the remaining capacity status of each module.

[0026] Step S200: Based on the operating status parameters and external control requirements, a control signal for controlling the on / off state of each switching unit is generated according to the topology control algorithm.

[0027] After acquiring operating status parameters and external control requirements, the central controller executes its built-in topology control algorithm for multi-objective optimization decision-making. First, the controller analyzes the external control requirements to determine the number of virtual energy storage units to be constructed and their respective performance indicators, including target output voltage levels, rated output power ranges, and preset operating modes. Next, based on real-time status data of the energy storage compartments, it filters out a set of available compartments whose state of charge is within the normal operating range, are in good health, and have no fault indicators. The topology control algorithm, aiming to match unit performance requirements and improve overall system efficiency, performs combination optimization calculations on the available compartments. By evaluating the output voltage matching degree under different electrical connection methods and the number of required power electronic conversion stages, it determines the optimal compartment combination scheme. Finally, based on the optimization results, a specific switching timing logic table is generated, which clearly defines the expected state and action time sequence of each switching unit.

[0028] Step S300: Based on the control signal, dynamically adjust the on / off state of the intelligent power electronic switch matrix to reconstruct the electrical connection relationship between several energy storage compartments.

[0029] After generating the control signals, the system enters the physical topology reconfiguration execution phase. The central controller processes the control signals according to their type: For fault isolation control signals, the controller identifies the fault compartment identifier, queries the topology connection mapping table, determines all switching units that need to be disconnected, generates an isolation execution command sequence, and controls the relevant switching units to operate synchronously, achieving multi-path electrical isolation of the fault compartment. For virtual unit construction control signals, the system parses the target topology connection diagram, generates a differentiated set of switching action commands, and executes the operation in stages according to the order of establishing series branches first and then parallel connections. Sufficient waiting intervals are introduced after each stage to ensure the safe recovery of power devices. Throughout the reconfiguration process, the system monitors the loop current and switching unit status in real time. By reading back the status register and measuring the insulation impedance parameters, it verifies the consistency between the actual electrical connection relationship and the expected target, ensuring the safe and reliable reconfiguration operation.

[0030] This invention achieves dynamic reconfigurability of the electrical connections in a compartmentalized energy storage system through the coordinated operation of an intelligent power electronic switch matrix and a topology control algorithm, significantly improving the system's operational flexibility and reliability. This method enables the system to quickly construct virtual energy storage units with specific electrical characteristics according to different application scenarios, while ensuring complete isolation of faulty compartments and continuous system operation. Through multi-dimensional optimization algorithms and a full-process safety verification mechanism, the overall energy efficiency of the system is improved while effectively ensuring the safety and reliability of the topology reconfiguration process, providing a complete technical solution for the efficient and stable operation of compartmentalized energy storage systems in complex power grid environments.

[0031] For example, consider a modular energy storage system with a rated power of 1MW / 2MWh. This system comprises 20 energy storage modules, each with a rated capacity of 100kW / 100kWh and an output voltage range of 400-800VDC. The intelligent power electronic switch matrix adopts a 5×8 array structure, containing 40 switching units. Each switching unit uses a 1700V / 400A IGBT module as the core switching device, equipped with corresponding drive and protection circuits. The outputs of the 20 energy storage modules are connected to 20 dedicated input ports of the switch matrix. When the system needs to participate in grid frequency regulation, the central controller, through topology control algorithms, can control the switch matrix to connect 10 modules in series to form an 800V high-voltage unit for rapid response to power regulation commands; simultaneously, it connects the other 10 modules in parallel to form a 400V high-capacity unit for continuous energy support. The two virtual energy storage units can operate independently, each catering to different grid service needs. When a fault occurs in an energy storage compartment, the central controller identifies the location of the faulty compartment using a fault detection algorithm, generates an isolation control command, and immediately disconnects all switching units connected to that compartment, achieving electrical isolation of the faulty compartment. Simultaneously, the system can automatically reconfigure the connection topology of the remaining 19 normal compartments, maintaining the system's continuous operation capability. The entire topology switching process is completed within milliseconds, ensuring the continuity and reliability of power grid services.

[0032] In one specific embodiment of the present invention, the state parameters include at least one of state of charge, health state, temperature and internal resistance, and the external control requirements include grid dispatch instructions, power allocation requirements or fault handling instructions.

[0033] Specifically, in step S200, based on operating state parameters and external control requirements, the central controller executes a topology control algorithm to generate control signals for controlling the on / off states of the intelligent power electronic switch matrix, including: Step S210: The central controller analyzes the external control requirements and determines the unit performance requirements of one or more virtual energy storage units to be formed, wherein the unit performance requirements include at least one of target output voltage, target output power and operating mode.

[0034] The central controller determines the required virtual energy storage unit configuration scheme by analyzing external control requests from the power grid dispatch center or energy management system. When the system receives a grid frequency regulation command, the controller identifies virtual energy storage units requiring high response rates. These units must have a target output voltage matching the grid interface voltage level, a target output power meeting the ramp-up requirements for frequency regulation, and operate in power point tracking mode. In peak-valley arbitrage scenarios, the controller analyzes load forecast data to identify large-capacity energy storage units requiring continuous discharge duration, and operates in energy throughput mode. By analyzing external commands in different application scenarios, the controller establishes a complete set of performance parameters for each virtual energy storage unit, including precise voltage levels, power ranges, and operating characteristics, providing clear objective function constraints for subsequent module combination optimization.

[0035] Step S220: Based on the operating status parameters, select available sub-compartments from several energy storage sub-compartments to form a set of available sub-compartments. The selection criteria include that the state of charge is in the effective working range, the health status is higher than the preset threshold, and there are no fault signs.

[0036] After determining the configuration requirements of the virtual energy storage units, the central controller performs availability assessment and screening of all energy storage modules based on real-time collected operating status parameters. First, it checks whether the state of charge (SOC) of each module is within the preset effective operating range. For virtual units about to participate in the discharge process, the SOC must be higher than the minimum discharge threshold; for units requiring charging, the SOC must be lower than the maximum charging threshold. Simultaneously, the controller evaluates the health status indicators of each module, excluding aging modules with health status below the preset threshold to ensure that modules participating in system operation have similar degradation characteristics. Furthermore, it checks the fault flags of each module, excluding modules with fault states such as abnormal temperature, abnormal voltage, or communication interruptions. Through these three screening criteria, a set of available modules that can be safely put into operation at the current moment is constructed, providing a reliable resource pool for topology optimization.

[0037] Step S230: With the goal of matching unit performance requirements and maximizing overall system efficiency, the energy storage compartments in the available compartment set are combined and optimized to calculate the optimal compartment combination scheme that satisfies the target electrical connection topology.

[0038] Based on the available compartment set and unit performance requirements, the central controller executes a multi-objective optimization algorithm to perform optimal compartment combination calculations. The optimization process prioritizes matching unit performance requirements while maximizing overall system efficiency. The controller iterates through all possible compartment combinations and their corresponding electrical connection topologies, evaluating the degree of matching between the output voltage of the virtual energy storage unit and the target output voltage under each topology. Simultaneously, it calculates the number of unnecessary power electronic conversion stages, such as DC-DC converters, required in the system's total loop. In frequency regulation applications, the optimization algorithm prioritizes series topologies that provide fast power response; while in energy-intensive applications, it tends to select parallel topologies that provide high-capacity output. The final output of the optimization algorithm is the compartment combination scheme with the highest overall system efficiency while satisfying all unit performance requirements. This scheme clearly defines the specific compartment members included in each virtual energy storage unit and their electrical connections.

[0039] Step S240: Generate the corresponding switching timing logic table according to the optimal compartment combination scheme.

[0040] After obtaining the optimal compartment combination scheme, the central controller converts it into directly executable switching control commands. Based on the electrical connection relationships defined in the compartment combination scheme, a corresponding switching timing logic table is generated. This logic table defines in detail the target state of each switching unit in the intelligent power electronic switching matrix, including the set of switching units that need to be closed and the set of switching units that need to be opened. Simultaneously, the logic table also specifies the precise timing of the switching unit state changes, ensuring that abnormal operating conditions such as loop short circuits or circulating currents do not occur during topology reconfiguration. For complex scenarios involving the simultaneous construction of multiple virtual energy storage units, the switching timing logic table adopts a phased execution strategy, prioritizing the establishment of basic connections for series branches, then completing the integration operation of parallel branches, and setting state verification points at key nodes to ensure the correct execution of each step.

[0041] Optionally, in one embodiment of the present invention, step S230, which involves optimizing the combination of energy storage compartments in the available compartment set to calculate the optimal compartment combination scheme that satisfies the target electrical connection topology, includes: Step S231a: Based on the unit performance requirements and the available compartment set, a candidate topology configuration set is generated, wherein each candidate topology configuration defines the series and parallel connection relationship of the energy storage compartments.

[0042] In the combinatorial optimization process, a complete set of candidate topology configurations is first generated based on the unit performance requirements and the available sub-module set. When constructing a virtual energy storage unit participating in the primary frequency regulation of the power grid, the controller generates candidate schemes for various topology configurations, including all-series, series-to-parallel, and parallel-to-series configurations, according to the target output voltage level and power requirements. For example, for a virtual unit with a target output voltage of 1500V and a power requirement of 500kW, the controller calculates how many sub-modules need to be connected in series to meet the voltage requirement, and how many such series branches need to be connected in parallel to meet the power requirement, thus generating all topology configurations that meet electrical constraints. When constructing large-capacity virtual units for peak shaving and valley filling, topology configurations with parallel connections are given priority to ensure the system's energy throughput capacity. Each candidate topology configuration clearly defines which sub-modules participate in the connection, what series and parallel connection methods are used, and the specific position of each sub-module in the topology, providing a complete comparative basis for subsequent quantitative evaluation.

[0043] Step S232a: For each candidate topology configuration in the candidate topology configuration set, calculate the performance consistency index of any virtual energy storage unit within the candidate topology configuration. The performance consistency index is calculated based on the variance of the state of charge and health status of all energy storage sub-compartments constituting the virtual energy storage unit.

[0044] For each candidate topology, the performance consistency of each virtual energy storage unit within it needs to be evaluated. In a series-connected virtual unit, the state-of-charge (POC) variance of all modules constituting the unit is calculated. If the POC differences among the modules of a unit are too large, some modules will reach their voltage limits before others during charging and discharging, thus affecting the overall usable capacity of the unit. Simultaneously, the variance of the health status of each module is calculated; modules with excessively large health status differences combined together will accelerate the overall aging process. For example, when constructing a virtual unit for grid frequency regulation, the POC variance of each module should not exceed 5%, and the health status variance should not exceed 3%, to ensure stable performance output during frequent charging and discharging. For parallel-connected virtual units, the consistency metrics are also evaluated, but the allowable variance range is relatively relaxed, focusing primarily on the current balancing characteristics of each parallel branch.

[0045] Step S233a: Calculate the system redundancy of the candidate topology configuration, wherein the system redundancy is quantified as follows: after simulating the failure and isolation of any energy storage compartment, the maximum number of independent combinations of virtual energy storage units that can be formed by topology reconstruction to meet the unit performance requirements in the remaining available energy storage compartments.

[0046] System redundancy assessment is a crucial step in topology optimization, measuring the continued service capability of candidate topology configurations under partial module failures. The assessment process involves simulating individual module failure scenarios. For each candidate topology configuration, the probability of rebuilding virtual energy storage units that meet the performance requirements of the original unit using the remaining available modules is simulated sequentially after each module fails and is isolated. For example, when assessing a virtual unit composed of ten modules, ten different failure scenarios are simulated, and the number of virtual units that can still meet the performance requirements under each scenario is calculated. The minimum value is taken as the system redundancy index for that candidate topology configuration. In grid-critical applications, the system requires a redundancy index of at least two to ensure that at least two independent virtual units can continue to provide service through topology reconfiguration even in the event of any single module failure.

[0047] Step S234a: Select a candidate topology configuration that simultaneously satisfies the performance consistency index being higher than the first preset threshold and the system redundancy index being higher than the second preset threshold, as the optimal compartment combination scheme.

[0048] After obtaining the performance consistency and system redundancy metrics for all candidate topologies, the final decision-making phase begins. The controller sets a first preset threshold to filter performance consistency, typically dynamically adjusted based on the application scenario of the virtual unit. A stricter threshold is set for power-type applications, while a more lenient threshold is set for energy-type applications. Simultaneously, a second preset threshold is set to filter system redundancy, adjusted according to the system's importance level, with critical application scenarios requiring a higher redundancy threshold. Topologies that simultaneously meet both threshold conditions are selected from the candidate set. If multiple configurations meet the conditions, the configuration with the higher performance consistency metric is prioritized; if none of the configurations simultaneously meet both thresholds, the performance consistency requirements are appropriately relaxed, prioritizing system redundancy. The final selected optimal compartment combination scheme ensures both the coordinated operation of each compartment within the virtual unit and the system's continuous service capability under fault conditions.

[0049] By employing a systematic process of candidate topology generation, performance consistency evaluation, system redundancy quantification, and multi-threshold screening, comprehensiveness and reliability of topology optimization for modular energy storage systems are achieved. This method not only considers performance optimization during normal system operation but also fully assesses system resilience under fault conditions, ensuring that the selected topology maintains excellent performance across various operating scenarios. Furthermore, the decision-making mechanism based on quantitative indicators avoids the subjectivity of human experience, improving the scientific rigor and repeatability of system decisions and providing a reliable technical foundation for the intelligent operation and maintenance of modular energy storage systems.

[0050] Furthermore, the calculation of the performance consistency index of any virtual energy storage unit within the candidate topology configuration in step S232a includes: Step S232a1: Obtain the real-time internal resistance and temperature parameters of all energy storage compartments constituting the virtual energy storage unit.

[0051] When evaluating the performance consistency of a virtual energy storage unit, the first step is to acquire real-time electrical and thermal parameters of all energy storage compartments that constitute the unit. Specifically, the DC internal resistance of each compartment is measured through the battery management system, reflecting the current conductivity and polarization characteristics of the battery. Simultaneously, real-time temperature data from multiple temperature measurement points within each compartment is collected, including cell surface temperature, bus connection point temperature, and radiator inlet and outlet temperatures. In grid frequency regulation applications, due to frequent power changes, these parameters are updated several times per second; while in peak shaving and valley filling scenarios, the update frequency can be appropriately reduced to several times per minute. Auxiliary parameters such as ambient temperature and cooling system operating status are also recorded, establishing a complete parameter monitoring system to provide accurate input data for subsequent power capacity calculations. Acquiring these real-time parameters ensures that the system can grasp the true state of each compartment under current operating conditions, avoiding the limitations of relying solely on historical data or nominal parameters for evaluation.

[0052] Step S232a2: Based on the state of charge, health status, real-time internal resistance and temperature parameters, calculate the maximum sustainable discharge power or maximum sustainable charging power of each energy storage compartment in the current operating mode.

[0053] Based on the acquired real-time parameters, the maximum sustainable power capacity of each energy storage module under current operating conditions is calculated. The calculation process comprehensively considers the constraints of state of charge on available capacity, the impact of health status on long-term performance, the limitations of real-time internal resistance on efficiency, and the constraints of temperature on safe operating boundaries. For modules in the discharge state, based on their current state of charge, internal resistance temperature rise characteristics, and heat dissipation conditions, the maximum discharge power that can be continuously maintained is calculated while ensuring that the voltage does not fall below the minimum cutoff voltage and the temperature does not exceed the safety threshold. Correspondingly, for modules in the charging state, the maximum charging power is calculated while ensuring that the voltage does not exceed the maximum charging voltage and the temperature rise is within a controllable range. When participating in grid frequency regulation, special attention is paid to the instantaneous power response capability of the modules, while in peak shaving and valley filling applications, more emphasis is placed on continuous power output characteristics. These calculations are based on the electrochemical characteristic curves and thermal management models of the modules to ensure that the obtained power values ​​are values ​​that can be safely and continuously executed in actual operation.

[0054] Step S232a3: The dispersion coefficient of the maximum sustainable discharge power or the maximum sustainable charging power of all energy storage compartments in the virtual energy storage unit is used as the dynamic power consistency index, where the dispersion coefficient is the ratio of the standard deviation to the mean.

[0055] After obtaining the maximum sustainable power data for each compartment, the power consistency level within the virtual energy storage unit is assessed. The coefficient of variation (COP) of the power values ​​for all compartments within the unit is calculated. This COP is obtained by comparing the standard deviation of the power values ​​to the average value, effectively eliminating the influence of dimensions and accurately reflecting the dispersion of power distribution. In a virtual unit composed of multiple compartments connected in series, if the COP of the maximum sustainable discharge power for each compartment is too large, the compartment with the weakest power capacity will reach its limit first during actual discharge, thus limiting the overall output level of the unit. For virtual units with parallel structures, an excessively large COP will lead to uneven current distribution in the parallel branches, affecting the overall efficiency and lifespan of the system. The COP threshold set by the system depends on the application scenario of the virtual unit. For frequency regulation applications requiring high consistency, the threshold is set more strictly; for energy-type applications, the threshold is relatively lenient but still needs to be controlled within a reasonable range.

[0056] Step S232a4 involves weighted fusion of the static consistency index and dynamic power consistency index calculated based on the variance of state of charge and health status to generate a comprehensive performance consistency index, wherein the weighting coefficients are dynamically adjusted according to the current operating mode of the virtual energy storage unit.

[0057] The final comprehensive performance consistency index is generated through weighted fusion, taking into full account the differences in the importance of various indicators under different operating modes. The static consistency index is calculated based on the variance of the state of charge and health states, reflecting the medium- to long-term state consistency of the modules; the dynamic power consistency index is based on the dispersion coefficient of power capacity, reflecting the real-time performance matching degree of the modules. In grid frequency regulation mode, a higher weight coefficient is assigned to the dynamic power consistency index because the power response speed and capacity matching of each module are crucial in this mode; in energy-type application mode, the weight of the static consistency index is appropriately increased, focusing more on the energy state balance of each module. The dynamic adjustment mechanism of the weight coefficients allows the system to optimize the evaluation criteria according to actual operating needs, ensuring that the most suitable topology configuration can be selected in different application scenarios. The final comprehensive performance consistency index is a normalized value between 0 and 1, with a higher value indicating better internal consistency of the virtual unit.

[0058] By establishing a comprehensive evaluation system that includes static state parameters and dynamic power parameters, a comprehensive and accurate evaluation of the performance consistency of virtual energy storage units was achieved. This evaluation method not only considers the long-term operating status of the compartments but also fully incorporates the impact of real-time operating conditions, enabling it to truly reflect the actual performance of the virtual units under various operating conditions. A dynamic weight adjustment mechanism based on operating modes ensures a close alignment between the evaluation results and actual application requirements, providing a scientific and reliable basis for topology optimization. This refined evaluation approach effectively improves the operating efficiency and reliability of virtual energy storage units, providing crucial technical support for the optimized operation of compartmentalized energy storage systems.

[0059] Optionally, in another embodiment of the present invention, step S230, which involves optimizing the combination of energy storage compartments in the available compartment set to calculate the optimal compartment combination scheme that satisfies the target electrical connection topology, includes: Step S231b: Obtain the current electrical connection topology of the compartmentalized energy storage system as the initial topology.

[0060] Before initiating the topology reconfiguration process, the current electrical connection status is first recorded and analyzed as the initial topology. By reading the real-time status of all switching units in the intelligent power electronic switch matrix and combining it with the connection relationship mapping table of each energy storage compartment, a complete current topology diagram is constructed. This diagram details which compartments are in series connection, which are in parallel connection, and the specific connection paths between each compartment. For example, when switching from grid frequency regulation mode to standby power mode, the initial topology may be a structure of multiple small-capacity series groups connected in parallel, while the target topology needs to be converted to a large-capacity parallel structure. By maintaining this real-time updated topology diagram, an accurate starting point is provided for subsequent reconfiguration path planning, while ensuring that abnormal states such as short circuits or open circuits can be effectively avoided during the reconfiguration process.

[0061] Step S232b: Based on the unit performance requirements and the available compartment set, generate a set of all feasible target topologies that can be reached from the initial topology through a single reconfiguration operation, where a single reconfiguration operation refers to a single topology switch achieved by changing the on / off state of the intelligent power electronic switch matrix.

[0062] Based on the initial topology and unit performance requirements, a set of all feasible target topologies is generated. This process is achieved by analyzing the state transition possibilities of each switch cell in the switch matrix, enumerating all new topologies that can be reached from the initial topology by changing only the state of a single switch cell. During generation, combinations of switch states that would lead to electrical connection conflicts or safety risks are first eliminated, such as switch groups that, when closed simultaneously, could form short-circuit loops. For each candidate target topology, it is verified whether it meets the voltage level and power capacity requirements in the unit performance requirements, while the electrical rationality of the topology structure is checked. For example, when constructing virtual cells with high voltage output, topologies that can form a sufficient number of compartments in series are selected; while in scenarios requiring high current output, topologies with more parallel connections are prioritized.

[0063] Step S233b: For each target topology in the feasible target topology set, calculate the reconstruction action cost required to get from it to the initial topology, where the reconstruction action cost is the weighted sum of the number of switch units that need to change state and the corresponding switch action time constant.

[0064] For each feasible target topology, the reconfiguration cost required to switch from the initial topology to the target topology is calculated. The quantification of the reconfiguration cost considers two main factors: the number of switching units that need to change state and the operating time characteristics of each switching unit. The total number of switches requiring a change in on / off state from the initial state to the target state is counted. Simultaneously, based on factors such as the mechanical characteristics, drive circuit response time, and arc-extinguishing capability of different switching units, a corresponding time constant weight is assigned to each switching unit. For example, fast semiconductor switches have smaller operating time constants, while high-capacity mechanical contactors have larger operating time constants. The total reconfiguration cost is calculated using a weighted summation method. This cost index comprehensively reflects the time and energy consumption required for topology switching, providing a quantitative basis for selecting the optimal reconfiguration path.

[0065] Step S234b: Select the target topology from the set of feasible target topologies that can meet the unit performance requirements and have the lowest reconstruction cost as the optimal compartment combination scheme.

[0066] After obtaining the reconfiguration costs of all feasible target topologies, the final decision-making phase begins. First, target topologies that fully meet the unit performance requirements are selected, including output voltage accuracy, power output capability, and operational stability. Then, among these qualified topologies, the scheme with the lowest reconfiguration cost is chosen as the optimal module combination scheme. During the selection process, both the speed and reliability of topology switching are considered, prioritizing topology switching paths involving fewer switching actions and shorter action times. For example, in grid emergency control scenarios, a scheme that can achieve topology reconfiguration with the fewest switching actions will be preferred to ensure rapid response to grid demands. The final selected optimal module combination scheme satisfies operational performance requirements while minimizing time and energy losses during the topology reconfiguration process.

[0067] By establishing a topology optimization mechanism based on the cost of reconfiguration actions, the efficiency of topology reconfiguration is maximized while meeting performance requirements. This optimization method fully considers the impact of actual equipment characteristics on the reconfiguration process, resulting in topology switching schemes that not only have superior electrical performance but also good engineering feasibility. The decision-making mechanism based on quantified costs ensures that the system can quickly select the optimal reconfiguration path, significantly improving the response speed and operational reliability of the compartmentalized energy storage system in dynamic operating environments, and providing the power grid with more flexible and efficient energy storage services.

[0068] Furthermore, step S232b, generating a set of all feasible target topologies reachable from the initial topology through a single reconstruction operation, includes: Step S232b1: Perform a safety prediction of the switching action on the initial topology, identify and eliminate all switching state combinations that would cause a short circuit or overcurrent in the power device, and obtain the safe switching action space.

[0069] In the initial stage of generating a feasible target topology set, a comprehensive safety analysis of the current topology is first performed. This process uses an electrical connection simulation model to predict whether all possible combinations of switching states will lead to system failure. A mathematical model of the topology connections is established based on Kirchhoff's voltage and current laws to analyze whether changing the state of any one or more switching units will create a short-circuit path in the loop. Simultaneously, based on the current voltage state and connection relationships of each energy storage compartment, potential instantaneous overcurrent situations are calculated, especially the circulating current impact that may occur when compartments of different voltage levels are directly connected in parallel. In grid frequency regulation applications, due to the need for frequent switching of operating states, the safety prediction process focuses on the transient processes under rapid switching actions to ensure that no electrical conflicts endangering equipment safety occur during dynamic reconfiguration. Through this proactive safety verification, a constrained switching action space is established, providing a safety guarantee for subsequent topology enumeration.

[0070] Step S232b2: Within the safety switch action space, enumerate all single switch state changes to generate the first layer transition topology state set.

[0071] After determining the safety switch action space, all possible single-step topology transformations are enumerated. This enumeration process is based on state transition theory in graph theory, treating the current topology as the initial state node and generating new topology state nodes by changing the state of individual switch units. All possible single-step state changes are systematically traversed according to the physical location and function of the switch units, including closing currently open switches to establish new connection paths, or opening currently closed switches to eliminate existing connections. During the enumeration process, the specific switch operation command corresponding to each state change and its expected topology change effect are recorded. For example, in a scenario transitioning from a parallel topology to a series topology, all possible first-step operations are enumerated, including opening certain parallel branch switches or closing certain series-connected switches, while ensuring that these operations are within the allowable range defined by the safety switch action space. This step generates all directly reachable transitional topology states from the initial topology.

[0072] Step S232b3: For each transition topology in the first layer transition topology state set, electrical parameters are verified based on unit performance requirements. Transition topologies whose output voltage deviation from the target output voltage is within the allowable range and whose loop current does not exceed the preset safety threshold are selected to obtain the second layer candidate topology set.

[0073] After obtaining the first-level transition topology state set, each transition topology undergoes rigorous electrical parameter verification. The verification process is based on the real-time state parameters of each energy storage compartment and the target performance requirements, calculating the system electrical characteristics under the transition topology through circuit simulation. First, the output voltage of the virtual energy storage unit is calculated to verify whether its deviation from the target output voltage is within the allowable range, typically requiring the deviation to not exceed five percent of the rated value. Simultaneously, based on the equivalent internal resistance and connection relationships of each compartment, the loop current distribution under rated power output conditions is calculated to ensure that the current in any branch does not exceed a preset safety threshold. In applications involving grid frequency regulation, the power response characteristics of the transition topology are also verified to ensure that it meets the special requirements of frequency regulation services for dynamic response. Only transition topologies that simultaneously meet voltage accuracy requirements, current safety constraints, and dynamic performance indicators are retained, forming the second-level candidate topology set.

[0074] In step S232b4, based on the second-layer candidate topology set, the single switch state change is enumerated again, and the safety prediction and electrical parameter verification are repeated until all generated topology states meet the unit performance requirements. The union of these topology states is taken as the feasible target topology set.

[0075] Building upon the second-layer candidate topology set, the topology space search continues to expand. This process employs an iterative deepening strategy, starting from the second-layer candidate topologies, enumerating single-switch state changes again to generate third-layer topology states, and repeatedly performing safety predictions and electrical parameter verifications. Each iteration expands the search depth while ensuring that newly generated topology states meet safety and performance requirements. Reasonable search depth limits are set to avoid combinatorial explosion while ensuring that reachable topologies meeting complex performance requirements are found. For example, when constructing a virtual energy storage unit with a specific voltage level and power capacity, multiple topology transformations may be required to reach the target state. This hierarchical and progressive search strategy systematically explores all possible reconfiguration paths. Finally, all verified topology states are merged to form a complete set of feasible target topologies, providing ample selection space for subsequent reconfiguration cost optimization.

[0076] By establishing a hierarchical and progressive topology generation and verification mechanism, feasible topology reconfiguration paths are comprehensively explored while ensuring safety. The advantage of this method lies in decomposing complex safety and performance verification into each search step, avoiding unsafe topology states from being considered later and ensuring that the final set of feasible topologies meets all system operational requirements. A screening process based on rigorous electrical parameter verification ensures the feasibility of each candidate topology for practical engineering implementation, while the iterative and in-depth search strategy ensures that no possible optimization solutions are overlooked, providing reliable technical support for the safe and efficient operation of compartmentalized energy storage systems.

[0077] Optionally, when the control signal includes a fault isolation control signal, step S300 involves dynamically adjusting the on / off state of the intelligent power electronic switch matrix based on the control signal to reconstruct the electrical connection relationship between several energy storage sub-compartments, including: Step S311: Extract the unique identifier of the faulty energy storage compartment to be isolated from the fault isolation control signal.

[0078] When a fault isolation command is detected or received, the central controller first parses and processes the fault isolation control signal. The controller extracts a unique identifier for the faulty energy storage compartment from the signal. This identifier is typically the compartment's physical address code or logical number within the system, ensuring accurate location of the specific faulty device. In grid frequency regulation applications, this process needs to be completed within a very short time, usually within ten milliseconds. Verifying the legality and validity of the identifier prevents incorrect isolation operations due to signal errors. Simultaneously, the controller records the timestamp of the fault occurrence and fault type information, providing data support for subsequent fault analysis and system maintenance. This step ensures that the system can accurately identify the target object requiring isolation, providing a clear target for subsequent isolation operations.

[0079] Step S312: Query the system topology connection relationship mapping table to determine the set of target switch units corresponding to all electrical connection paths with the faulty energy storage compartment.

[0080] After identifying the faulty compartment to be isolated, a pre-stored topology connection mapping table is queried to analyze all electrical connection paths of the faulty compartment within the current topology. This mapping table records in detail the connection relationships between each energy storage compartment and each switch unit in the switch matrix, including all direct and indirect connections. A topology analysis algorithm identifies all switch units electrically connected to the faulty compartment, regardless of whether these switch units are in a closed or open state. In complex hybrid topologies, this analysis process needs to consider series branches, parallel connections, and possible cross connections to ensure that no potential electrical connection path is overlooked. The analysis results form a target switch unit set, which contains all switch units required to achieve complete electrical isolation of the faulty compartment.

[0081] Step S313: Generate a fault isolation execution instruction sequence. The fault isolation execution instruction sequence controls all switch units in the target switch unit set to synchronously switch to the open state, thereby realizing multi-path electrical isolation of the faulty energy storage compartment.

[0082] Based on the target set of switching units, a specific fault isolation execution command sequence is generated. This command sequence clearly specifies the operation (transition to the open state) to be performed by each target switching unit, the timing of the operation, and the timing relationship between each operation. To ensure thoroughness and safety of isolation, a synchronous transition strategy is adopted, using precise timing control to ensure that all target switching units complete the state transition simultaneously within a very short time window. This synchronous operation avoids transient overvoltage or arcing problems that may be caused by sequential disconnection. The command sequence also includes necessary delay settings and state confirmation steps to ensure that subsequent operations are executed only after the previous operation is completed. When generating the command sequence, the electrical characteristics of the switching units, such as breaking capacity and tripping capability, are also considered to ensure that the isolation operation will not damage the switching equipment.

[0083] Step S314: After the preset isolation verification time window, the status register of the target switch unit set and the insulation impedance parameters of the system's total circuit are read back to verify that the faulty energy storage compartment has been completely isolated and there is no residual leakage current.

[0084] After completing the isolation operation, the system enters the isolation effect verification phase. The system waits for a preset isolation verification time window, which takes into account the arc extinguishing time required for the switchgear to completely disconnect and the decay time of the system transient process. The verification process employs a dual verification mechanism: first, by reading back the status register of each switch unit in the target switch unit set, it is confirmed that all switches are indeed in the open state; simultaneously, the insulation impedance parameters of the system's total circuit are measured to verify whether the insulation performance of the system has returned to normal after the fault compartment isolation. In critical power grid applications, the system will also monitor the system stability after isolation to ensure that the isolation operation has not affected the normally operating parts. Only when all verification indicators meet the preset requirements is it confirmed that the fault compartment has been completely isolated; otherwise, the backup isolation scheme will be activated or an alarm signal will be issued.

[0085] By establishing a complete fault isolation process, rapid, accurate, and reliable isolation of faulty energy storage compartments was achieved. This mechanism ensures that faulty equipment can be promptly isolated from the system in the event of a fault, preventing the fault from spreading and affecting the normal operation of other parts of the system. The multi-path electrical isolation strategy fundamentally severs all electrical connections between the faulty compartment and the system, while the dual verification mechanism guarantees the reliability and integrity of the isolation operation. This fault handling capability significantly improves the reliability and safety of the compartmentalized energy storage system, providing an important guarantee for the stable operation of the system in the power grid.

[0086] Optionally, when the control signal includes a virtual unit control signal, step S300 involves dynamically adjusting the on / off state of the intelligent power electronic switch matrix based on the control signal to reconstruct the electrical connection relationship between several energy storage compartments, including: Step S321: Analyze the target topology connection diagram contained in the virtual unit control signal. The target topology connection diagram defines the connection relationship between several target energy storage sub-compartments in the form of a node adjacency matrix.

[0087] Upon receiving the virtual unit control signal, the central controller first parses the target topology connection diagram contained within it. This connection diagram is expressed mathematically as a node adjacency matrix, where the rows and columns of the matrix correspond to the energy storage sub-module nodes in the system, and the values ​​of the matrix elements represent the connection relationships between the corresponding sub-modules. When constructing a virtual energy storage unit participating in the primary frequency regulation of the power grid, the target topology connection diagram may define a specific structure consisting of eight sub-modules first connected in pairs to form four series groups, and then these four series groups are connected in parallel. The parsing process includes verifying the completeness and rationality of the node adjacency matrix, checking for isolated nodes or illegal connection relationships, and confirming that the connection relationships defined in the diagram can be physically implemented through a switch matrix. It also calculates the expected electrical parameters based on the target topology connection diagram, including total output voltage, total output power, and equivalent internal resistance, providing a reference benchmark for subsequent verification stages.

[0088] Step S322: Compare the target topology connection diagram with the current topology state of the system to generate a differentiated set of switching action instructions. The set of switching action instructions includes the switching units whose states need to be changed and their target states.

[0089] The parsed target topology connection diagram is precisely compared with the current actual topology state to generate a differentiated set of switching action instructions. The comparison process is achieved by comparing the expected and actual states of each switching unit under the target and current states, generating operation instructions only for those switching units with inconsistent states. For example, in a scenario where the current topology is a fully parallel structure while the target topology is a series-to-parallel structure, the parallel connections that need to be disconnected and the series connections that need to be closed are identified, forming a specific list of switching actions. When generating the instruction set, the dependencies and timing requirements of the switching actions are considered to ensure that the execution order of the instructions complies with electrical safety principles. Simultaneously, the instruction sequence is optimized to minimize unnecessary switching operations, reduce equipment wear, and improve reconfiguration efficiency. This optimization is particularly important in power grid frequency regulation applications that require frequent topology switching.

[0090] Step S323: The switching action instruction set is executed in stages according to the order of establishing the series branch first and then the parallel connection, and a waiting interval of no less than the power device recovery time is introduced after each stage operation is completed.

[0091] When executing the switching action command set, a phased sequential execution strategy is adopted, strictly following the order of establishing series branches first, followed by establishing parallel connections. During the series branch construction phase, the corresponding switching units are controlled to close, forming independent series links, and state verification is performed after each link is established. After each phase operation is completed, a waiting interval of no less than the power device recovery time is introduced. This interval is determined based on the characteristics of the switching equipment used; for mechanical contactors, it typically requires tens of milliseconds, while for solid-state switches, it only requires a few microseconds. The waiting interval ensures that the switching devices reach a stable operating state after state switching, avoiding malfunctions caused by residual charge or magnetic flux changes. The parallel connection phase only begins after all series branches have been established. This phased execution strategy effectively prevents abnormal situations such as circulating currents or short circuits during topology reconfiguration.

[0092] Step S324: After all switching actions have been completed, the voltage at the connection point and the direction of the loop current of each target energy storage compartment are sampled to verify the consistency between the actual electrical connection relationship and the target topology connection diagram.

[0093] After all switching actions are completed, the topology verification phase begins. The verification process involves sampling the voltage values ​​at each target energy storage compartment connection point and the current direction in the loops. The measured voltage relationships are compared with the expected voltage relationships calculated based on the target topology connection diagram. For example, in a series structure, it is verified whether the total voltage equals the sum of the voltages of each compartment; in a parallel structure, it is verified whether the voltages of each parallel branch are equal. Simultaneously, Hall effect current sensors are used to detect the current direction in each loop to verify whether the current path conforms to expectations. In grid frequency regulation applications, this verification process needs to be completed quickly, typically requiring verification results within hundreds of milliseconds. If an inconsistency between the actual connection relationship and the target topology is detected, the specific discrepancy information is recorded, and corresponding corrective measures or alarm procedures are initiated to ensure the system operates under the correct topology.

[0094] By establishing a complete control process from topology analysis to execution verification, the construction process of virtual energy storage units is made more precise and reliable. The topology description method based on the node adjacency matrix provides a clear and unambiguous definition of connection relationships; the differentiated switching action command generation mechanism improves the efficiency of reconfiguration operations; the phased execution strategy ensures the safety of the reconfiguration process; and the electrical parameter verification stage ensures the accuracy of the reconfiguration results. This complete implementation scheme enables the modular energy storage system to quickly and reliably construct virtual energy storage units that meet various application requirements, significantly improving the system's adaptability and practicality in complex power grid environments.

[0095] In addition, after generating the corresponding switching timing logic table based on the optimal compartment combination scheme in step S250, the following steps are also included: Step S261: Configure an independent operation control loop for at least one virtual energy storage unit, wherein each operation control loop calculates and executes the charging and discharging power command of the virtual energy storage unit based on the unit performance requirements of its corresponding virtual energy storage unit.

[0096] After completing the topology construction of the virtual energy storage units, an independent operation control loop is configured for each virtual unit. These control loops are individually tuned based on the specific performance requirements of their respective virtual units. For example, virtual units participating in grid frequency regulation use a fast power point tracking algorithm, while those used for peak shaving and valley filling employ an energy dispatch algorithm. Each control loop continuously monitors the internal state of its assigned virtual unit, including remaining capacity, power output capability, and operating efficiency, while simultaneously receiving power commands from the upper-level energy management system. Based on the real-time state and performance requirements of the virtual unit, the control loop calculates precise charging and discharging power commands using proportional-integral-derivative (PID) control algorithms or more advanced model predictive control (MMC) algorithms. These commands not only consider current power demands but also take into account long-term operational optimization of the virtual unit, ensuring that grid service requirements are met while extending equipment lifespan.

[0097] Step S262: Monitor the voltage and frequency fluctuations of the bus connecting each virtual energy storage unit in real time. When the fluctuations exceed the preset stable range, dynamically allocate damping power commands based on the remaining regulation capacity of at least one virtual energy storage unit.

[0098] High-precision sensors monitor the voltage and frequency parameters of the bus connected to each virtual energy storage unit in real time, with sampling frequencies typically reaching thousands of times per second to meet grid quality control requirements. When a bus voltage deviation exceeds two percent of the rated value or a frequency deviation exceeds ±0.1 Hz, the grid is determined to be in an unstable state, and a damping control mechanism is activated. First, the real-time regulation capability of each virtual energy storage unit is evaluated, including available charge / discharge power margin, response speed, and duration. Based on the evaluation results, damping power commands are dynamically allocated according to a preset optimization objective function. Typically, virtual units with fast response speeds and high regulation accuracy are prioritized for high-frequency, small-amplitude regulation, while virtual units with large capacity and high inertia are assigned to continuous power support tasks.

[0099] Step S263: The damping power command is superimposed with the original charging and discharging power command to generate the final power command of at least one virtual energy storage unit, and executed through the operation control loop to achieve coordinated and stable control of the system bus by at least one virtual energy storage unit.

[0100] After determining the damping power command for each virtual unit, it is vector-superimposed with the original basic power command to generate the final power execution command. The superposition process must consider the priority and timing characteristics of different power commands to ensure that damping adjustment does not affect the basic function execution of the virtual unit. The final power command is decomposed into specific switching control signals through a control loop, driving the AC / DC converters of each virtual unit to execute the corresponding power output. In scenarios with grid frequency fluctuations, multiple virtual units coordinate control to jointly provide inertia support and primary frequency regulation services, forming a cooperative stabilization effect. The effect of the cooperative control is monitored in real time, and the power allocation ratio of each virtual unit is dynamically adjusted through a closed-loop feedback mechanism to ensure that the bus voltage and frequency parameters quickly recover and stabilize within acceptable ranges.

[0101] By establishing independent control and collaborative stabilization mechanisms for virtual energy storage units, multi-timescale power support capabilities are achieved under complex grid operating conditions. Independent operation control loops ensure that each virtual unit can be finely controlled according to its own characteristics and task requirements, while the real-time state-based damped power allocation mechanism fully leverages the overall system's regulation potential. Intelligent superposition and coordinated execution of power commands enable the system to simultaneously meet multiple grid service demands, significantly enhancing the practical value and operational reliability of compartmentalized energy storage systems in the grid. This hierarchical and collaborative control architecture provides an effective technical solution for large-scale energy storage systems to participate in grid ancillary services.

[0102] Furthermore, after acquiring the real-time operating status parameters of several energy storage sub-compartments in step S100, and before generating the control signal for controlling the on / off state of each switching unit according to the topology control algorithm based on the operating status parameters and external control requirements in step S200, the method further includes: Step S110: Based on historical temperature data and real-time internal resistance data in the operating status parameters, predict the internal resistance growth trend and capacity decay trajectory of each energy storage compartment in the future preset period through an electrochemical aging model.

[0103] After acquiring the operating status parameters of each energy storage module, an electrochemical aging prediction and analysis process is initiated. This process, based on long-term collected historical temperature data and real-time measured internal resistance data, combines an electrochemical aging model to predict the future performance degradation of each module. First, trend analysis is performed on the historical temperature data to identify typical thermal cycling patterns experienced by the modules during operation, including key parameters such as high-temperature operating duration, temperature fluctuation amplitude, and frequency. Simultaneously, the changing patterns of real-time internal resistance measurements are analyzed to establish a correlation model between internal resistance and the number of cycles and operating temperature. Based on this data, the electrochemical aging model calculates the expected trajectory of internal resistance growth and the capacity degradation curve for each module over the next one hundred charge-discharge cycles or thirty days of operation. In grid frequency regulation applications, due to frequent power changes, the focus is on the accelerating effect of high-frequency charging and discharging on module aging; while in peak shaving and valley filling scenarios, the long-term impact of deep charging and discharging on capacity degradation is more important.

[0104] Step S120: Identify energy storage compartments whose internal resistance growth trend slope or capacity decay rate exceeds the adaptive early warning threshold and mark them as potential risk compartments.

[0105] After obtaining the performance degradation prediction data for each module, potential risk identification and analysis are performed. The slope of the internal resistance growth trend for each module is calculated, representing the expected increase in internal resistance per unit time, and compared with an adaptive warning threshold determined based on the module type and operating history. Simultaneously, the capacity degradation rate is assessed, calculating the percentage of capacity loss per unit cycle. The adaptive warning threshold is not a fixed value but is dynamically adjusted based on the module's operating time, cumulative cycle count, and operating environment. For example, a stricter internal resistance growth threshold is used for modules that have been in operation for more than three years; for modules primarily involved in grid frequency regulation, the warning threshold for the capacity degradation rate is correspondingly relaxed. When the slope of the internal resistance growth trend exceeds 0.1 milliohms per hour or the capacity degradation rate exceeds 0.5 percent per 100 cycles, it is marked as a potentially risky module, and the specific risk type and severity level are recorded.

[0106] In step S130, when executing the topology control algorithm in the future, an operational constraint strategy is applied to the potential risk compartment. The operational constraint strategy includes limiting its maximum charging and discharging current, avoiding placing it in a virtual energy storage unit with high power demand, or reducing its priority in combinatorial optimization.

[0107] For identified potentially risky modules, differentiated operation and management strategies are implemented in subsequent topology control algorithms. For modules with excessively rapid internal resistance growth, current limits are imposed on the virtual energy storage units they participate in, controlling the maximum charge and discharge current to below 80% of the rated value to reduce further aging caused by Joule heating. For modules with significant capacity degradation, their priority is reduced during the combinatorial optimization process to avoid configuring them in virtual energy storage units with high capacity consistency requirements. When constructing high-power virtual units for grid frequency regulation, potentially risky modules are prioritized for exclusion to ensure the reliability of frequency regulation services; however, when constructing virtual units for energy transfer, these modules can be used to a limited extent, but enhanced operational monitoring is required. The charge and discharge cutoff voltages of these modules are also dynamically adjusted according to the risk level, appropriately narrowing the operating voltage range to extend their service life.

[0108] By establishing a predictive maintenance mechanism based on an electrochemical aging model, early identification and preventative intervention of performance degradation in energy storage modules were achieved. This mechanism transforms traditional passive fault handling into proactive performance management. By considering the aging trends of modules in advance during the topology control phase, it effectively slows down the equipment performance degradation process and significantly improves the long-term operational reliability of the system. Differentiated operational constraint strategies maximize the service life of high-risk modules while ensuring overall system performance, optimizing the operational economy throughout the entire lifecycle. This predictive maintenance capability provides crucial technical support for the intelligent operation and maintenance of modular energy storage systems, and is particularly valuable in grid application scenarios requiring high reliability.

[0109] Accordingly, a second aspect of the present invention provides an electronic device, including: at least one processor and a memory connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, cause the at least one processor to perform the aforementioned compartmentalized energy storage system control method.

[0110] Accordingly, a third aspect of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-described compartmentalized energy storage system control method.

[0111] The embodiments of the present invention aim to protect a control method for a compartmentalized energy storage system, which has the following effects: 1. By deploying an intelligent power electronic switch matrix and executing topology control algorithms, the system completely breaks free from the constraints of fixed electrical connections, enabling dynamic reconfiguration of the connection relationships of each energy storage compartment within milliseconds. This not only allows the system to "fission" into multiple virtual energy storage units with different voltages and power ratings according to grid demand, simultaneously undertaking different application tasks, but also achieves complete physical isolation of faulty compartments, fundamentally eliminating fault propagation paths and solving the key problems of rigid operating modes and reliability limitations imposed by fixed topologies in existing systems; 2. Beyond traditional simple power allocation, the optimization process simultaneously considers static performance balance, dynamic power capability, system redundancy, and the cost and safety of the reconfiguration process itself. This multi-objective and multi-constraint optimization ensures that the topology selected by the system under any operating condition is not only the most efficient at present, but also the most resilient to future failures, and the safest and most economical for the switching process. This significantly extends the service life of the equipment while improving the overall efficiency of the system. 3. It not only includes accurate and rapid isolation and verification processes after a fault and closed-loop safety verification during the construction of virtual units, but also innovatively introduces predictive maintenance strategies based on electrochemical models and control mechanisms for the coordinated stabilization of the power grid by multiple virtual units. This forms a complete technical closed loop from cell-level potential risk warning to network-level reconfiguration safety execution, and then to system-level proactive support, elevating system safety management from "passive response to faults" to a new level of intelligence of "proactive prediction, proactive defense, and proactive support".

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method of controlling a compartmentalized energy storage system, the method comprising: The modular energy storage system comprises a central controller, a plurality of energy storage modules and an intelligent power electronic switch matrix connected between the plurality of energy storage modules, the intelligent power electronic switch matrix comprising a plurality of switch unit arrays, the central controller controlling the on-off state of each switch unit in the intelligent power electronic switch matrix, and the control method is applied to the central controller and comprises the following steps: Real-time acquisition of operating state parameters and external control requirements of the plurality of energy storage modules; Based on the operating state parameters and the external control requirements, a control signal for controlling the on-off state of each switch unit is generated according to a topology control algorithm; According to the control signal, the on-off state of the intelligent power electronic switch matrix is dynamically adjusted, and the electrical connection relationship between the plurality of energy storage modules is reconstructed.

2. The compartmentalized energy storage system control method of claim 1, wherein, The state parameters include at least one of state of charge, state of health, temperature and internal resistance, and the external control requirements include grid dispatching instructions, power distribution requirements or fault handling instructions; Based on the operating state parameters and the external control requirements, a control signal for controlling the on-off state of the intelligent power electronic switch matrix is generated according to a topology control algorithm, which comprises: Analyzing the external control requirements to determine the unit performance requirements of one or more virtual energy storage units to be formed, wherein the unit performance requirements include at least one of target output voltage, target output power and operating mode; Based on the operating state parameters, available modules are selected from the plurality of energy storage modules to form an available module set, wherein the selection conditions include that the state of charge is in the effective working interval, the state of health is higher than the preset threshold and there is no fault flag; Combination optimization is performed on the energy storage modules in the available module set to calculate an optimal module combination scheme that meets the target electrical connection topology, with the optimization objectives being matching the unit performance requirements and maximizing the overall system efficiency; According to the optimal module combination scheme, a corresponding switch timing logic table is generated.

3. The compartmentalized energy storage system control method of claim 2, wherein, The combination optimization is performed on the energy storage modules in the available module set to calculate an optimal module combination scheme that meets the target electrical connection topology, which comprises: Based on the unit performance requirements and the available module set, a candidate topology configuration set is generated, wherein each candidate topology configuration defines the series and parallel connection relationship of energy storage modules; For each candidate topology configuration in the candidate topology configuration set, a performance consistency index of any virtual energy storage unit in the candidate topology configuration is calculated, the performance consistency index being calculated based on the variance of the state of charge and the state of health of all energy storage modules constituting the virtual energy storage unit; The system redundancy of the candidate topology configuration is calculated, wherein the system redundancy is quantified by simulating that any energy storage module fails and is isolated, and then the maximum number of independent combinations of virtual energy storage units that meet the unit performance requirements can be formed in the remaining available energy storage modules through topology reconstruction. Select the candidate topology configuration which satisfies the performance consistency index higher than a first preset threshold and the system redundancy index higher than a second preset threshold as the optimal compartment combination scheme.

4. The compartmentalized energy storage system control method of claim 3, wherein, The performance consistency index of any virtual energy storage unit in the candidate topology configuration is calculated, including: Obtaining real-time internal resistance and temperature parameters of all energy storage compartments constituting the virtual energy storage unit; Based on the state of charge, state of health, real-time internal resistance and temperature parameters, the maximum sustainable discharge power or the maximum sustainable charge power of each energy storage compartment under the current operating mode is calculated; The discrete coefficient of the maximum sustainable discharge power or the maximum sustainable charge power of all energy storage compartments in the virtual energy storage unit is taken as the dynamic power consistency index, wherein the discrete coefficient is the ratio of the standard deviation to the average value; The static consistency index calculated based on the state of charge and state of health variance is weighted and fused with the dynamic power consistency index to generate a comprehensive performance consistency index, wherein the weight coefficient is dynamically adjusted according to the current operating mode of the virtual energy storage unit.

5. The compartmentalized energy storage system control method of claim 2, wherein, The energy storage compartments in the available compartment set are combined and optimized to calculate the optimal compartment combination scheme that satisfies the target electrical connection topology, including: Obtaining the current electrical connection topology of the compartment type energy storage system as the initial topology; Based on the unit performance demand and the available compartment set, a set of all feasible target topologies that can be reached from the initial topology through a one-time reconstruction operation is generated, wherein a one-time reconstruction operation refers to a single topology switching achieved by changing the on-off state of the intelligent power electronic switch matrix; For each target topology in the set of feasible target topologies, the reconstruction action cost required from it to the initial topology is calculated, wherein the reconstruction action cost is the weighted sum of the number of switch units that need to change state and the corresponding switching action time constant; From the set of feasible target topologies, select the target topology that can satisfy the unit performance demand and has the minimum reconstruction action cost as the optimal compartment combination scheme.

6. The compartmentalized energy storage system control method of claim 5, wherein, The set of all feasible target topologies that can be reached from the initial topology through a one-time reconstruction operation is generated, including: Performing switch action safety prediction on the initial topology to identify and exclude all switch state combinations that will cause loop short circuit or power device overcurrent to obtain a safe switch action space; Enumerate all single switch state changes in the safe switch action space to generate a first layer of transition topology state set; For each transition topology in the first layer of transition topology state set, perform electrical parameter checking based on the unit performance demand to filter out transition topologies whose deviation between output voltage and target output voltage is within the allowed range and whose loop current does not exceed the preset safety threshold to obtain a second layer of candidate topology set; On the basis of the second layer of candidate topology set, single switch state changes are enumerated again, and safety prediction and electrical parameter checking are repeated until all generated topology states satisfy the unit performance demand, and the union of these topology states is taken as the set of feasible target topologies.

7. The compartmentalized energy storage system control method of claim 2, wherein, The control signal comprises a fault isolation control signal; The dynamic adjustment of the on-off state of the intelligent power electronic switch matrix according to the control signal to reconstruct the electrical connection relationship between the plurality of energy storage compartments comprises: Resolving the unique identifier of the fault energy storage compartment to be isolated from the fault isolation control signal; Querying a system topology connection relationship mapping table to determine a corresponding target switch unit set on all electrical connection paths of the fault energy storage compartment; Generating a fault isolation execution instruction sequence, which controls all switch units in the target switch unit set to be synchronously converted to an open state, thereby realizing multi-path electrical isolation of the fault energy storage compartment; After a preset isolation verification time window, verifying that the fault energy storage compartment has been completely isolated and has no residual leakage current by reading back the state registers of the target switch unit set and the system total loop insulation impedance parameters.

8. The compartmentalized energy storage system control method of claim 2, wherein, The control signal comprises a virtual unit control signal; The dynamic adjustment of the on-off state of the intelligent power electronic switch matrix according to the control signal to reconstruct the electrical connection relationship between the plurality of energy storage compartments comprises: Resolving a target topology connection graph contained in the virtual unit control signal, the target topology connection graph defining the connection relationship between a plurality of target energy storage compartments in the form of a node adjacency matrix; Comparing the target topology connection graph with the current system topology state to generate a differential switch action instruction set, the switch action instruction set containing switch units that need to change state and their target states; According to the order of first establishing a series branch and then establishing a parallel connection, the switch action instruction set is executed in stages, and a waiting interval of no less than the power device recovery time is introduced after each stage of operation is completed; After all switch actions are executed, the consistency of the actual formed electrical connection relationship with the target topology connection graph is verified by sampling the connection point voltage and loop current direction of each target energy storage compartment.

9. The control method of claim 2-8, wherein, After generating the corresponding switch timing logic table according to the optimal compartment combination scheme, the method further comprises: Configuring independent operation control loops for the at least one virtual energy storage unit, wherein each operation control loop calculates and executes the charge and discharge power instructions of the corresponding virtual energy storage unit based on the unit performance requirements of the virtual energy storage unit; Real-time monitoring of the voltage and frequency fluctuations of each virtual energy storage unit connection bus, when the fluctuations exceed the preset stable interval, dynamically allocating damping power instructions based on the remaining regulation capacity of at least one virtual energy storage unit; Superimposing the damping power instructions and the original charge and discharge power instructions to generate the final power instructions of at least one virtual energy storage unit, and executing the final power instructions through the operation control loop to realize the cooperative stability control of at least one virtual energy storage unit on the system bus.

10. The compartmentalized energy storage system control method of claim 9, wherein, After real-time acquisition of the operating state parameters of the plurality of energy storage compartments, before generating the control signal for controlling the on-off state of each switch unit based on the operating state parameters and the external control demand according to the topology control algorithm, the method further comprises: Based on historical temperature data and real-time internal resistance data in the operating state parameters, the internal resistance growth trend and capacity decay trajectory of each energy storage sub-cabin in a future preset period are predicted through an electrochemical aging model; Identify the energy storage sub-cabin whose internal resistance growth trend slope or capacity decay rate exceeds the adaptive warning threshold, and mark it as a potential risk sub-cabin; When executing the topology control algorithm subsequently, an operation restriction strategy is applied to the potential risk sub-cabin, which includes limiting its maximum charge and discharge current, avoiding placing it in a virtual energy storage unit with high power demand, or reducing its priority in combined optimization.

Images