Network-constructed energy storage coordination control method, system, device and medium

Abstract

This invention relates to the field of power system technology and discloses a method, system, device, and medium for coordinated control of grid-connected energy storage. The method constructs a dynamic admittance matrix using the mixed topology parameters between the star-shaped main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located. Based on the voltage measurement data of the common connection point of the grid-connected energy storage system and the dynamic admittance matrix, it identifies whether the grid-connected energy storage system has entered a fault condition and optimizes the fractional-order VSG control parameters of the grid-connected energy storage system. It also performs high-precision state estimation of the observed state data of each energy storage unit using an improved multi-rate Runge-Kutta method. Furthermore, based on the pre-acquired total output power of the grid-connected energy storage system, it optimizes power allocation according to the state estimation values ​​of each energy storage unit, thereby improving the adaptability and operational reliability of the grid-connected energy storage system under complex operating conditions.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, system, device and medium for coordinated control of grid-connected energy storage. Background Technology

[0002] With the high proportion of new energy sources being connected to the grid, transient stability control of grid-forming energy storage systems (GFM-ESS) has become a core challenge for grid upgrades. Traditional VSG (Virtual Synchronous Generator) solutions, due to their fixed integer order model and centralized optimization, face problems such as insufficient Virtual Short-Circuit Ratio, power angle instability, and response delay.

[0003] However, existing control strategies for grid-type energy storage are mostly limited to a single time scale, making it difficult to effectively address dynamic coupling issues across multiple time and space scales. In particular, they are limited in complex operating conditions where high-frequency fluctuations and low-frequency energy dispatch coexist. Summary of the Invention

[0004] In view of this, in order to solve the above-mentioned technical problems, the present invention provides a method, system, device and medium for coordinated control of grid-connected energy storage.

[0005] The first aspect of this invention provides a method for coordinated control of grid-connected energy storage, comprising:

[0006] Based on the hybrid topology parameters between the star-shaped main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located, a dynamic admittance matrix is ​​constructed; wherein, the dynamic admittance matrix is ​​used to characterize the impedance coupling relationship between the star-shaped main grid and the honeycomb units under multiple spatiotemporal scales;

[0007] The voltage measurement data of the common connection point of the grid-connected energy storage system is obtained, and the fault flag bit of the grid-connected energy storage system is determined by combining the dynamic admittance matrix; wherein, the fault flag bit is used to characterize the dual fault characteristics of grid voltage drop and current over-limit.

[0008] Based on reinforcement learning algorithms, the fractional-order VSG control parameters of the grid-connected energy storage system are adjusted in conjunction with the fault flag bits.

[0009] The observed state data of each energy storage unit in the grid-connected energy storage system and the fractional-order VSG control parameters are obtained. The state estimation of the observed state data is performed by combining the improved multi-rate Runge-Kutta method to obtain the state estimation value of each energy storage unit.

[0010] Based on the total output power of the grid-connected energy storage system obtained in advance, power allocation optimization is performed according to the state estimation values ​​of each energy storage unit, power commands for each energy storage unit are generated, and the power commands for each energy storage unit are sent to the corresponding energy storage units for execution.

[0011] Preferably, the hybrid topology parameters include the line impedance of the star-shaped main grid, the real-time virtual impedance of the grid-connected energy storage system, the filter inductance, and the grid angular frequency;

[0012] Based on the hybrid topology parameters between the star-shaped main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located, a dynamic admittance matrix is ​​constructed, including:

[0013] Obtain the topological connection structure between the star main grid and the honeycomb unit within the power grid topology where the grid-connected energy storage system is located, and identify the physical coupling relationship between the star main grid and the honeycomb unit;

[0014] Based on the real-time virtual impedance, filter inductance, and grid angular frequency of the grid-connected energy storage system, the time-varying admittance matrix of the honeycomb unit is constructed.

[0015] Based on the line impedance of the star-shaped main network, determine the real admittance matrix of the star-shaped main network.

[0016] Based on the physical coupling relationship between the star-shaped main grid and the honeycomb unit, and combining the direct admittance from the star-shaped main grid to the honeycomb unit and the time-varying admittance matrix of the honeycomb unit, the coupling admittance matrix between the star-shaped main grid and the honeycomb unit is determined; wherein, the direct admittance is the complex admittance of the physical branch connection between the star-shaped main grid and the honeycomb unit;

[0017] The time-varying admittance matrix of the honeycomb unit, the real admittance matrix of the star-shaped main grid, and the coupling admittance matrix are integrated in the form of a block matrix to synthesize the dynamic admittance matrix.

[0018] Preferably, the step of obtaining voltage measurement data at the point of common coupling of the grid-connected energy storage system and determining the fault flag bit of the grid-connected energy storage system in conjunction with the dynamic admittance matrix includes:

[0019] Obtain voltage measurement data at the common connection point of the grid-connected energy storage system;

[0020] Based on the voltage measurement data and the dynamic admittance matrix, calculate the injection current vector at the common connection point of the grid-connected energy storage system;

[0021] Based on the voltage measurement data and the injected current vector, it is determined whether the grid-connected energy storage system has experienced both voltage drop and current over-limit faults.

[0022] If it is determined that the grid-connected energy storage system has experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 1;

[0023] If it is determined that the grid-connected energy storage system has not experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 0.

[0024] Preferably, the fractional-order VSG control parameters include the fractional-order order;

[0025] The step of adjusting the fractional-order VSG control parameters of the grid-connected energy storage system based on the reinforcement learning algorithm and in conjunction with the fault flag bits includes:

[0026] The state space and action space of the reinforcement learning algorithm are determined based on the fractional order and the order adjustment amount.

[0027] Based on the fault flag, and in conjunction with the voltage deviation and short-circuit ratio of the grid-connected energy storage system, the reward function is determined;

[0028] Based on the reward function, and combining the state space and the action space, the fractional order is updated using the policy gradient method until the convergence condition is met. The latest fractional order is then output as the adjusted fractional VSG control parameter.

[0029] Preferably, the energy storage unit includes a lithium battery subsystem and a supercapacitor subsystem; the state estimate includes the state of charge estimate of the lithium battery subsystem and the voltage estimate of the supercapacitor subsystem.

[0030] The process of acquiring the observed state data of each energy storage unit in the grid-connected energy storage system and the fractional-order VSG control parameters, and then using the improved multi-rate Runge-Kutta method to perform state estimation on the observed state data to obtain the state estimate value of each energy storage unit includes:

[0031] Acquire observational status data of the lithium battery subsystem and the supercapacitor subsystem; wherein, the observational status data includes the state-of-charge rate of the lithium battery subsystem and the port voltage of the supercapacitor subsystem;

[0032] Based on the rate of change of state of charge of the lithium battery subsystem, a third-order Runge-Kutta explicit iterative scheme is constructed based on the implicit correction mechanism, and implicit correction coefficients and adaptive step size are introduced to predict the estimated state of charge of the lithium battery subsystem.

[0033] The fractional derivative term of the supercapacitor subsystem is determined based on the port voltage of the supercapacitor subsystem and the fractional-order VSG control parameters.

[0034] Based on the fractional derivative terms of the supercapacitor subsystem, a fourth-order Runge-Kutta implicit iterative scheme is constructed using an explicit correction mechanism. Implicit correction coefficients and adaptive step sizes are introduced to predict the voltage estimate of the supercapacitor subsystem.

[0035] Preferably, the step of optimizing power allocation based on the pre-acquired total output power of the grid-connected energy storage system according to the state estimate of each energy storage unit, generating power commands for each energy storage unit, and issuing the power commands of each energy storage unit to the corresponding energy storage unit for execution includes:

[0036] Based on the response characteristics of the supercapacitor subsystem and the lithium battery subsystem, the high-frequency power component of the total output power is allocated to the supercapacitor subsystem, and the low-frequency power component is allocated to the lithium battery subsystem; wherein, the high-frequency power component and the low-frequency power component are obtained by performing fractional-order differential filtering on the total output power;

[0037] Based on the total output power and the response time constant of the supercapacitor subsystem, the theoretical value of the power distribution of the supercapacitor subsystem is determined;

[0038] Based on the voltage estimate of the supercapacitor subsystem, the safe power limit of the supercapacitor subsystem is determined, and the minimum value between the theoretical power allocation value and the safe power limit is taken as the power allocation value of the supercapacitor subsystem.

[0039] Based on the total output power and the theoretical power allocation value of the supercapacitor subsystem, the remaining power is determined as the theoretical power allocation value of the lithium battery subsystem;

[0040] Based on the theoretical power allocation value of the lithium battery subsystem and the estimated state of charge value of the lithium battery subsystem, the power allocation value of the lithium battery subsystem is determined.

[0041] Based on the power allocation values ​​of the supercapacitor subsystem and the lithium battery subsystem, power commands for the supercapacitor subsystem and the lithium battery subsystem are generated respectively.

[0042] The power commands of the supercapacitor subsystem and the lithium battery subsystem are respectively sent to the supercapacitor subsystem and the lithium battery subsystem for execution.

[0043] Preferably, the method further includes:

[0044] The initial VSG virtual impedance of the grid-connected energy storage system is obtained, and the initial VSG virtual impedance is corrected by a fractional order to obtain the VSG virtual impedance after fractional order correction.

[0045] Based on the fractionally corrected VSG virtual impedance and the equivalent capacitance of the supercapacitor subsystem, the impedance partial derivative of the supercapacitor subsystem is determined.

[0046] Based on the fractionally corrected VSG virtual impedance, combined with the equivalent internal resistance and nominal capacity of the lithium battery subsystem, the impedance bias of the lithium battery subsystem is determined.

[0047] The Jacobian matrix is ​​constructed using the impedance partial derivatives of the supercapacitor subsystem and the lithium battery subsystem.

[0048] Based on the Jacobian matrix, the maximum Lyapunov exponent is determined, and the stability of the grid-connected energy storage system is evaluated based on the maximum Lyapunov exponent.

[0049] Secondly, the present invention also provides a grid-based energy storage collaborative control system, comprising:

[0050] The dynamic admittance construction module is used to construct a dynamic admittance matrix based on the mixed topology parameters between the star main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located; wherein, the dynamic admittance matrix is ​​used to characterize the impedance coupling relationship between the star main grid and the honeycomb units at multiple spatiotemporal scales;

[0051] The fault flag determination module is used to acquire voltage measurement data of the common connection point of the grid-connected energy storage system, and determine the fault flag bit of the grid-connected energy storage system in combination with the dynamic admittance matrix; wherein, the fault flag bit is used to characterize the dual fault characteristics of grid voltage drop and current over-limit.

[0052] The control parameter optimization module is used to adjust the fractional-order VSG control parameters of the grid-connected energy storage system based on a reinforcement learning algorithm and in conjunction with the fault flag bit.

[0053] The state estimation module is used to acquire the observed state data of each energy storage unit of the grid-connected energy storage system and the fractional-order VSG control parameters, and to perform state estimation on the observed state data by combining the improved multi-rate Runge-Kutta method to obtain the state estimation value of each energy storage unit.

[0054] The power allocation execution module is used to optimize the power allocation according to the state estimate of each energy storage unit based on the total output power of the grid-connected energy storage system obtained in advance, generate power instructions for each energy storage unit, and send the power instructions of each energy storage unit to the corresponding energy storage unit for execution.

[0055] Thirdly, the present invention also provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, the computer program being executed by the processor causing the processor to perform the steps of the grid-based energy storage coordinated control method as described in the first aspect.

[0056] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the steps of the grid-based energy storage coordinated control method as described in the first aspect.

[0057] As can be seen from the above technical solutions, this invention constructs a dynamic admittance matrix by using the hybrid topology parameters between the star-shaped main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located. This allows for real-time evolution of the impedance coupling relationship between the star-shaped main grid and the honeycomb units at multiple temporal and spatial scales. Based on the voltage measurement data of the common connection point of the grid-connected energy storage system and the dynamic admittance matrix, it identifies whether the grid-connected energy storage system has entered a fault condition and optimizes and adjusts the fractional-order VSG control parameters of the grid-connected energy storage system. This ensures that the VSG of the grid-connected energy storage system takes into account both fault response speed and steady-state support accuracy. Furthermore, it improves the multi-rate Runge-Kutta method to perform high-precision state estimation of the observed state data of each energy storage unit, ensuring the coupling consistency of the observed state data of each energy storage unit at multiple time scales. Based on the pre-acquired total output power of the grid-connected energy storage system, it optimizes power allocation according to the state estimation values ​​of each energy storage unit, realizing coordinated response and power allocation of each energy storage unit at multiple temporal and spatial scales, and improving the adaptability and operational reliability of the grid-connected energy storage system under complex operating conditions. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is an application environment diagram of a grid-based energy storage collaborative control method provided in an embodiment of the present invention;

[0060] Figure 2 A flowchart of a grid-based energy storage collaborative control method provided in an embodiment of the present invention;

[0061] Figure 3 This is a schematic diagram of a grid-connected energy storage collaborative control system provided in an embodiment of the present invention;

[0062] Figure 4This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0063] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] The grid-based energy storage coordinated control method provided in this application embodiment can be applied to, for example... Figure 1 The application environment shown is illustrated. Terminal 101 communicates with server 102 via a network. A data storage system can store the data that server 102 needs to process. The data storage system can be integrated onto server 102, or it can be located in the cloud or on another network server. Terminal 101 or server 102 constructs a dynamic admittance matrix based on the hybrid topology parameters between the star-shaped main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located. The dynamic admittance matrix characterizes the impedance coupling relationship between the star-shaped main grid and the honeycomb units at multiple spatiotemporal scales. Voltage measurement data of the common coupling point of the grid-connected energy storage system is acquired, and the fault flag bits of the grid-connected energy storage system are determined in conjunction with the dynamic admittance matrix. The fault flag bits characterize the dual fault characteristics of grid voltage sag and current overrun. Based on a reinforcement learning algorithm, the fractional-order VSG control parameters of the grid-connected energy storage system are adjusted in conjunction with the fault flag bits. The observed state data and fractional-order VSG control parameters of each energy storage unit in the grid-connected energy storage system are acquired, and the state estimation of the observed state data is performed using the improved multi-rate Runge-Kutta method to obtain the state estimate value of each energy storage unit. Based on the pre-acquired total output power of the grid-connected energy storage system, power allocation optimization is performed according to the state estimate values ​​of each energy storage unit, generating power commands for each energy storage unit, and the power commands of each energy storage unit are sent to the corresponding energy storage units for execution.

[0065] Terminal 101 can be, but is not limited to, various personal computers, laptops, smartphones, and tablets.

[0066] Server 102 can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides cloud computing services.

[0067] like Figure 2 As shown in the embodiments of this application, a method for coordinated control of grid-connected energy storage is provided, which is applied to... Figure 1Taking terminal 101 or server 102 as an example, the explanation includes the following steps S1 to S5. Wherein:

[0068] Step S1: Construct a dynamic admittance matrix based on the hybrid topology parameters between the star main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located; wherein, the dynamic admittance matrix is ​​used to characterize the impedance coupling relationship between the star main grid and the honeycomb units at multiple temporal and spatial scales.

[0069] The star-shaped main grid and the honeycomb units are the power grid topologies upon which the grid-connected energy storage system relies for coordinated control. Together, they form a honeycomb-star hybrid topology. The star-shaped main grid is the upper-level backbone of the entire system, typically using substations or core switching stations as star sources and ring main units and switching stations as main nodes, forming a radial star connection. The star-shaped main grid carries the low-frequency, high-capacity power exchange of the entire system (such as the second-level response power of lithium batteries), and also serves as the connection carrier for the honeycomb units, dynamically interacting with the lower-level honeycomb units. It is the main channel for the grid-connected energy storage system to access the power grid.

[0070] A honeycomb cell is a distributed subnet cell under a star-shaped main network. A single honeycomb cell is a regular hexagonal subnet. Its high-frequency nonlinear characteristics are characterized by fractional calculus, which complements the low-frequency response of the star-shaped main network cell in a spatiotemporal manner.

[0071] The honeycomb-star hybrid topology organizes dispersed grid-connected energy storage units (such as lithium batteries and supercapacitors in different locations) according to the logic of upper-layer backbone and lower-layer distribution. It achieves multi-temporal and spatial scale power collaborative allocation through a dynamic admittance matrix, which maps the time-varying impedance characteristics between the star main grid and the honeycomb units in real time.

[0072] Step S2: Obtain the voltage measurement data of the common coupling point of the grid-connected energy storage system, and determine the fault flag bit of the grid-connected energy storage system in combination with the dynamic admittance matrix; wherein, the fault flag bit is used to characterize the dual fault characteristics of grid voltage drop and current over-limit.

[0073] The voltage measurement data of the point of common coupling (PCC) of the grid-connected energy storage system is obtained by high-frequency sampling. The voltage measurement data and dynamic admittance matrix are used to determine whether the grid-connected energy storage system is in a dual fault state of voltage drop and current over-limit. Generally speaking, when the voltage amplitude of the PCC is lower than 80% of the rated value and the injected current exceeds the set threshold, the fault flag bit is triggered and the protection mechanism is started.

[0074] Step S3: Based on the reinforcement learning algorithm, adjust the fractional-order VSG control parameters of the grid-connected energy storage system in combination with the fault flag bit.

[0075] Fractional-order VSG control parameters are the core control parameters of the VSG in a grid-connected energy storage system, including fractional-order differential orders. Through reinforcement learning algorithms, the fractional-order VSG control parameters are used as the agent's state, and fault flags are used as reward signals to optimize the fractional-order VSG control parameters. This rapidly improves the system's transient support capability under fault conditions, balancing dynamic response and steady-state accuracy.

[0076] Step S4: Obtain the observed state data and fractional-order VSG control parameters of each energy storage unit in the grid-connected energy storage system. Combine the improved multi-rate Runge-Kutta method to perform state estimation on the observed state data and obtain the state estimate value of each energy storage unit.

[0077] Among them, the observed state data are multi-source heterogeneous measurement data such as real-time voltage, current, state of charge (SOC) and temperature of each energy storage unit. The improved multi-rate Runge-Kutta method integrates the time-varying characteristics of fractional-order VSG control parameters with the time scale difference of the observed state data, and uses the improved multi-rate Runge-Kutta solver to perform high-precision state estimation on the observed state data to obtain the state estimate value of each energy storage unit.

[0078] Step S5: Based on the total output power of the grid-connected energy storage system obtained in advance, optimize the power allocation according to the state estimation value of each energy storage unit, generate power commands for each energy storage unit, and send the power commands of each energy storage unit to the corresponding energy storage units for execution.

[0079] The total output power is issued based on control commands, and the power commands of each energy storage unit are dynamically allocated according to the state estimate of each energy storage unit and its dynamic response characteristics. Generally speaking, supercapacitors prioritize responding to high-frequency power demands, while lithium batteries bear the base load and low-frequency fluctuations, ensuring the power balance and energy optimization of the system at multiple time and space scales.

[0080] It should be noted that the embodiments of this application construct a dynamic admittance matrix by using the hybrid topology parameters between the star main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located. This allows for real-time evolution of the impedance coupling relationship between the star main grid and the honeycomb units at multiple temporal and spatial scales. Based on the voltage measurement data of the common connection point of the grid-connected energy storage system and the dynamic admittance matrix, it identifies whether the grid-connected energy storage system has entered a fault condition and optimizes and adjusts the fractional-order VSG control parameters of the grid-connected energy storage system. This ensures that the VSG of the grid-connected energy storage system takes into account both fault response speed and steady-state support accuracy. Furthermore, it improves the multi-rate Runge-Kutta method to perform high-precision state estimation of the observed state data of each energy storage unit, ensuring the coupling consistency of the observed state data of each energy storage unit at multiple time scales. Additionally, based on the pre-acquired total output power of the grid-connected energy storage system, it optimizes power allocation according to the state estimation values ​​of each energy storage unit, realizing coordinated response and power allocation of each energy storage unit at multiple temporal and spatial scales, and improving the adaptability and operational reliability of the grid-connected energy storage system under complex operating conditions.

[0081] In some embodiments, the hybrid topology parameters include the line impedance of the star-shaped main grid, the real-time virtual impedance of the grid-connected energy storage system, the filter inductance, and the grid angular frequency. In this case, a dynamic admittance matrix is ​​constructed based on the hybrid topology parameters between the star-shaped main grid and the honeycomb units within the grid topology where the grid-connected energy storage system is located, including the following steps S101~S105:

[0082] Step S101: Obtain the topological connection structure between the star main grid and the honeycomb unit within the power grid topology where the grid-connected energy storage system is located, and identify the physical coupling relationship between the star main grid and the honeycomb unit.

[0083] Among them, the star-shaped main grid is the upper-level backbone grid of the entire system, and the honeycomb unit is the distributed sub-grid unit under the star-shaped main grid. The topological connection between the star-shaped main grid and the honeycomb unit organizes the dispersed grid-connected energy storage units (such as lithium batteries and supercapacitors in different locations) according to the logic of upper-level backbone and lower-level distribution.

[0084] Step S102: Construct the time-varying admittance matrix of the honeycomb unit based on the real-time virtual impedance, filter inductance and grid angular frequency of the grid-connected energy storage system.

[0085] Among them, through real-time virtual impedance Filter inductor Grid angular frequency Determine the equivalent impedance of the honeycomb cell Where j is the imaginary unit, the equivalent impedance of the honeycomb unit is taken as its reciprocal and multiplied by the three-phase conversion factor 3 (adapted to three-phase AC power grid) and the impedance adjustment factor. (range of values) ), to obtain the time-varying admittance matrix of a single honeycomb unit, that is:

[0086]

[0087] In the formula, This is the time-varying admittance of the honeycomb unit.

[0088] Step S103: Determine the real admittance matrix of the star-shaped main grid based on the line impedance of the star-shaped main grid.

[0089] Specifically, by diagonalizing the reciprocal of the real part of the line impedance of the star-shaped main network, the real admittance matrix of the star-shaped main network is formed, namely:

[0090]

[0091] In the formula, The real admittance of the star-shaped main grid structure.

[0092] Step S104: Based on the physical coupling relationship between the star main grid and the honeycomb unit, and combining the direct admittance from the star main grid to the honeycomb unit and the time-varying admittance matrix of the honeycomb unit, determine the coupling admittance matrix between the star main grid and the honeycomb unit; wherein, the direct admittance is the complex admittance of the physical branch connection between the star main grid and the honeycomb unit.

[0093] Among them, the direct admittance from the star-shaped main grid to the honeycomb unit This is the real part of the branch admittance between the star main grid and the honeycomb unit. By eliminating intermediate redundant nodes using the Kron reduction method, the coupling admittance matrix between the star main grid and the honeycomb unit is obtained, i.e.:

[0094]

[0095] In the formula, This is the coupling admittance matrix.

[0096] Step S105: Integrate the time-varying admittance matrix of the honeycomb unit, the real admittance matrix of the star main grid, and the coupled admittance matrix in the form of a block matrix to synthesize the dynamic admittance matrix.

[0097] After obtaining the time-varying admittance matrix of the honeycomb unit, the real admittance matrix of the star main network, and the coupling admittance matrix, they are integrated in the form of block matrices. At the same time, they are extended to the entire network dimension through matrix operations to ensure coverage of the dynamic response of all nodes.

[0098] Dynamic admittance matrix It is a block-symmetric matrix with the following structure:

[0099]

[0100] In the formula, The Kronecker product is used to extend the honeycomb unit admittance to the entire network dimension. Expand to matrix, It is a 6th-order identity matrix used for block diagonalization of the honeycomb cell admittance, ensuring that each honeycomb cell is calculated independently.

[0101] In some embodiments, obtaining voltage measurement data at the point of common coupling of the grid-connected energy storage system and determining the fault flag bit of the grid-connected energy storage system in conjunction with the dynamic admittance matrix includes the following steps S201~S205:

[0102] Step S201: Obtain voltage measurement data of the common connection point of the grid-connected energy storage system.

[0103] The voltage measurement data is obtained in real time by smart meters or phasor measurement units (PMUs).

[0104] Step S202: Calculate the injection current vector at the common connection point of the grid-connected energy storage system based on the voltage measurement data and the dynamic admittance matrix.

[0105] The product of the voltage measurement data and the dynamic admittance matrix is ​​the node injection current vector, i.e. ,in, This is a voltage measurement.

[0106] Step S203: Based on the voltage measurement data and the injected current vector, determine whether the grid-connected energy storage system has experienced both voltage drop and current over-limit faults.

[0107] Among them, by setting the maximum allowable injection current and nominal voltage (Grid rated parameters (e.g., 380V low-voltage system or 35kV medium-voltage system)), using voltage measurement data and injected current vector respectively compared with nominal voltage and maximum allowable injection current A comparison is made to achieve dual verification of fault criteria; when the voltage norm at the point of common coupling is lower than 80% of the nominal voltage. Furthermore, the node injection current norm exceeds the maximum allowable injection current. If the voltage drops and current exceedances occur simultaneously, the grid-connected energy storage system is considered to have experienced a dual fault; otherwise, the system is considered to be operating normally. The judgment rule is as follows:

[0108]

[0109] In the formula, This is a fault flag (Boolean value).

[0110] Step S204: If it is determined that the grid-connected energy storage system has a dual fault of voltage drop and current over-limit, then set the fault flag bit to 1.

[0111] The fault flag bit being 1 indicates that the system is in a fault condition.

[0112] Step S205: If it is determined that the grid-connected energy storage system has not experienced both voltage drop and current over-limit faults, then set the fault flag bit to 0.

[0113] The fault flag being 0 indicates that the system is in normal operating condition.

[0114] In some embodiments, the fractional-order VSG control parameters include a fractional order; in this case, adjusting the fractional-order VSG control parameters of the grid-connected energy storage system based on a reinforcement learning algorithm and in conjunction with fault flag bits includes the following steps S301~S303:

[0115] Step S301: Determine the state space and action space of the reinforcement learning algorithm based on the fractional order and order adjustment amount.

[0116] Fractional order is used to characterize response sensitivity. The higher the order, the stronger the damping response of the system to voltage disturbances, but the dynamic response is lagging. The lower the order, the faster the transient tracking, but the more likely it is to cause high-frequency oscillations.

[0117] This application aims to dynamically adjust the fractional order, using the fractional order as an action variable in reinforcement learning and the order adjustment as the action space. The fractional order is then updated using the order adjustment via the policy gradient. .

[0118] Step S302: Determine the reward function based on the fault flag bit, combined with the voltage deviation and short-circuit ratio of the grid-connected energy storage system.

[0119] The reinforcement learning strategy is dynamically switched through the fault flag: when Fault_flag=1, an emergency control strategy with high reward weights is activated, prioritizing the suppression of overcurrent and improving voltage support capability. The reward function focuses on voltage recovery, i.e., the reward function is set as follows:

[0120]

[0121] In the formula, As a reward value, For voltage deviation, This is the short-circuit ratio.

[0122] When Fault_flag=0, the reward rules focus on grid strength, i.e., the reward function is set as follows:

[0123]

[0124] Step S303: Based on the reward function, and combining the state space and action space, update the fractional order using the policy gradient method until the convergence condition is met, and output the latest fractional order as the adjusted fractional VSG control parameter.

[0125] The fractional order is updated through policy gradient. Through the reward function Dynamically adjust the fractional order of the gradient direction To balance response speed and stability, the update is as follows:

[0126]

[0127] In the formula, For the updated fractional order, The policy gradient learning rate, (t is the number of training steps), T is the total number of steps. A discount factor (typically 0.9) is used to reinforce learning.

[0128] Updated The fractional differential equations acting immediately on the VSG change the system's dynamic response. The adjustment process terminates when any of the following conditions are met:

[0129] Gradient norm (Tend to stability) or Fluctuation range less than 1% (e.g., within 10 consecutive steps) ), where k is the current iteration step.

[0130] Once the termination condition is met, the updated fractional order is output as the adjusted fractional order VSG control parameter.

[0131] In some embodiments, the energy storage unit includes a lithium battery subsystem and a supercapacitor subsystem; the state estimate includes the state of charge estimate of the lithium battery subsystem and the voltage estimate of the supercapacitor subsystem; in this case, the observed state data and fractional-order VSG control parameters of each energy storage unit in the grid-connected energy storage system are obtained, and the state estimate of each energy storage unit is obtained by combining the improved multi-rate Runge-Kutta method with the observed state data, including the following steps S401~S404:

[0132] Step S401: Obtain the observation status data of the lithium battery subsystem and the supercapacitor subsystem; wherein, the observation status data includes the rate of change of state of charge of the lithium battery subsystem and the port voltage of the supercapacitor subsystem.

[0133] Specifically, for the lithium battery subsystem, the current of the lithium battery subsystem is collected. Rated capacity and conversion efficiency The rate of change of state of charge of the lithium battery subsystem is calculated using the ampere-hour integration method, i.e.:

[0134]

[0135] In the formula, This represents the rate of change of the state of charge of the lithium battery subsystem.

[0136] Step S402: Based on the rate of change of state of charge of the lithium battery subsystem, a third-order Runge-Kutta explicit iterative scheme is constructed based on the implicit correction mechanism, and implicit correction coefficients and adaptive step size are introduced to predict the estimated state of charge of the lithium battery subsystem.

[0137] The implicit correction mechanism utilizes explicit RK3 to suppress high-frequency oscillations by introducing a frequency domain filtering term. Specifically, the third-order Runge-Kutta explicit iterative scheme is as follows:

[0138]

[0139]

[0140]

[0141]

[0142] In the formula, , , The term represents the intermediate slope. It reflects the instantaneous impact of the current on the SOC. Correcting errors caused by time-varying current characteristics Suppress high-frequency noise, This is the right-hand function of the dynamic equation for a lithium battery. For the current moment, This is the low-frequency integration step size (adaptive step size). , These are implicit correction coefficients. , The time constant of the lithium battery. This represents the state of charge (SOC) of the lithium battery at step n. This is the estimated state of charge (SOC) value for the lithium battery at step n+1. For the second derivative term in the state space.

[0143] Step S403: Determine the fractional derivative term of the supercapacitor subsystem based on the port voltage and fractional-order VSG control parameters of the supercapacitor subsystem.

[0144] Among them, the fractional-order VSG control parameters include the fractional-order differential order, based on the updated fractional-order order. This only affects the calculation of the fractional derivative of the supercapacitor and does not participate in the RK3 integration of the lithium battery. Using the fractional derivative order and the port voltage of the supercapacitor subsystem, the fractional derivative term of the supercapacitor subsystem is determined as follows:

[0145]

[0146] In the formula, As a fractional derivative term, it can accurately characterize the high-frequency voltage fluctuations of supercapacitors at the millisecond level. For gamma function, Where N is the sampling time interval, N is the number of historical moments, and k is the sequence number of the historical moment.

[0147] Step S404: Based on the fractional derivative terms of the supercapacitor subsystem, a fourth-order Runge-Kutta implicit iterative scheme is constructed based on an explicit correction mechanism, and implicit correction coefficients and adaptive step size are introduced to predict the voltage estimate of the supercapacitor subsystem.

[0148] Among them, the fourth-order Runge-Kutta implicit iterative scheme is constructed using the explicit correction mechanism RK4 as follows:

[0149]

[0150]

[0151]

[0152]

[0153]

[0154] In the formula, This represents the current time in milliseconds for the supercapacitor. The high-frequency integration step size (value h1) (satisfies Nyquist's theorem) The fourth slope of RK4, () is the formula for the voltage change rate of a supercapacitor. For the current moment The terminal voltage of a supercapacitor The weighting coefficient is the rate of change. Here, λ is the weighting coefficient of RK4, λ is the frequency domain filter strength factor, and ξ is the frequency domain filter coefficient: , The cutoff frequency, Let be the transfer function of the high-frequency noise filter. This is the estimated value of the supercapacitor terminal voltage at the next moment after filtering correction.

[0155] In some embodiments, based on the pre-acquired total output power of the grid-connected energy storage system, power allocation optimization is performed according to the state estimation values ​​of each energy storage unit, generating power commands for each energy storage unit, and sending the power commands of each energy storage unit to the corresponding energy storage units for execution, including the following steps S501~S507:

[0156] Step S501: Based on the response characteristics of the supercapacitor subsystem and the lithium battery subsystem, the high-frequency power component of the total output power is allocated to the supercapacitor subsystem, and the low-frequency power component is allocated to the lithium battery subsystem; wherein, the high-frequency power component and the low-frequency power component are obtained by fractional-order differential filtering of the total output power.

[0157] Among them, due to the respective response characteristics of the supercapacitor subsystem and the lithium battery subsystem, the supercapacitor subsystem has a relatively fast response speed and is specifically designed to cope with sudden power fluctuations in the power grid (such as a sudden stop in photovoltaic power generation or a sudden increase in load), but its endurance is poor and it is suitable for high-frequency power components. The supercapacitor subsystem responds to millisecond-level power changes, smooths out voltage spikes, and avoids system instability; while the lithium battery subsystem focuses on responding to low-frequency power demands at the second level or above, and undertakes the main energy support.

[0158] In this process, by performing fractional-order differential filtering on the total output power, high-frequency power components and low-frequency power components are obtained, thus achieving spatiotemporal decoupling of power.

[0159] Step S502: Determine the theoretical power distribution value of the supercapacitor subsystem based on the total output power and the response time constant of the supercapacitor subsystem.

[0160] The total output power is the virtual synchronous machine active power command calculated based on control commands and the grid angular frequency, namely:

[0161]

[0162] In the formula, Total output power This is the active power command (i.e., electromagnetic torque) for the virtual synchronous machine. ,in, The equivalent rotational inertia of VSG, It is a fractional-order angular acceleration. For angular frequency deviation, is the damping coefficient.

[0163] The theoretical power distribution value of the supercapacitor subsystem is determined by the total output power and the response time constant of the supercapacitor subsystem:

[0164]

[0165] In the formula, This represents the theoretical value for power distribution in the supercapacitor subsystem. For the response time constant ( ).

[0166] Step S503: Based on the voltage estimate of the supercapacitor subsystem, determine the safe power limit of the supercapacitor subsystem, and take the minimum value between the theoretical power allocation value and the safe power limit as the power allocation value of the supercapacitor subsystem.

[0167] The safe power limit of the supercapacitor subsystem is calculated as follows:

[0168]

[0169] In the formula, The safe power limit for the supercapacitor subsystem. This refers to the capacitance of the supercapacitor.

[0170] To avoid supercapacitors Since the energy released internally must not exceed its safety limit, the minimum value between the theoretical power allocation value and the safe power limit is taken as the power allocation value of the supercapacitor subsystem, that is:

[0171]

[0172] In the formula, This refers to the power allocation value for the supercapacitor subsystem.

[0173] Step S504: Based on the total output power and the theoretical power allocation value of the supercapacitor subsystem, determine the remaining power as the theoretical power allocation value of the lithium battery subsystem.

[0174] Specifically, the remaining power within the total output power, excluding the theoretical power allocation value of the supercapacitor subsystem, should be the theoretical power allocation value of the lithium battery subsystem, i.e.:

[0175]

[0176] In the formula, This represents the theoretical value for power distribution in the lithium battery subsystem.

[0177] Step S505: Determine the power allocation value of the lithium battery subsystem based on the theoretical power allocation value of the lithium battery subsystem and the estimated state of charge value of the lithium battery subsystem.

[0178] To avoid overcharging and over-discharging of lithium batteries, the actual power of the lithium battery subsystem is limited based on the estimated state of charge (SOC) value.

[0179]

[0180] In the formula, This refers to the power allocation value for the lithium battery subsystem.

[0181] Step S506: Generate power commands for the supercapacitor subsystem and the lithium battery subsystem respectively based on the power allocation values ​​of the supercapacitor subsystem and the lithium battery subsystem.

[0182] Among them, the power command is .

[0183] Step S507: Send the power command of the supercapacitor subsystem and the power command of the lithium battery subsystem to the supercapacitor subsystem and the lithium battery subsystem respectively for execution.

[0184] In some embodiments, to evaluate the stability of executing power commands, the method further includes: obtaining the initial VSG virtual impedance of the grid-connected energy storage system, and performing a fractional-order correction on the initial VSG virtual impedance to obtain the fractionally corrected VSG virtual impedance; determining the impedance partial derivative of the supercapacitor subsystem based on the fractionally corrected VSG virtual impedance and the equivalent capacitance of the supercapacitor subsystem; determining the impedance partial derivative of the lithium battery subsystem based on the fractionally corrected VSG virtual impedance and the equivalent internal resistance and nominal capacity of the lithium battery subsystem; constructing a Jacobian matrix using the impedance partial derivatives of the supercapacitor subsystem and the lithium battery subsystem; determining the maximum Lyapunov exponent based on the Jacobian matrix, and evaluating the stability of the grid-connected energy storage system based on the maximum Lyapunov exponent.

[0185] The initial VSG virtual impedance is calculated as follows:

[0186]

[0187] In the formula, Let be the time-varying admittance of the honeycomb cell at the initial time t0. As a reference virtual impedance, For the operation of taking the imaginary part; where:

[0188]

[0189] In the formula, This is the rated output voltage of the inverter. Rated capacity for VSG.

[0190] The initial VSG virtual impedance is corrected by a fractional order to obtain the corrected VSG virtual impedance as follows:

[0191]

[0192] In the formula, The VSG virtual impedance is the fractional-order corrected impedance. Adjust the gain for virtual impedance (empirical value) ), To modify the activation function of the linear unit, make sure It only increases during high-frequency disturbances to avoid losses under normal operating conditions. The frequency derivative term is estimated using a fractional-order Kalman filter to suppress noise interference.

[0193] The impedance partial derivative of the supercapacitor subsystem is:

[0194]

[0195] In the formula, This represents the impedance partial derivative of the supercapacitor subsystem.

[0196] The impedance partial derivative of the lithium battery subsystem is:

[0197]

[0198] In the formula, For the impedance partial derivative of the lithium battery subsystem, This is the equivalent internal resistance of the lithium battery subsystem. This refers to the nominal capacity.

[0199] The Jacobian matrix is:

[0200]

[0201] In the formula, It is a Jacobian matrix.

[0202] Calculating the maximum Lyapunov exponent using the Jacobian matrix for:

[0203]

[0204] Comparison of the largest Lyapunov index and The size of the largest Lyapunov exponent Less than If the value is positive, it indicates that the grid-connected energy storage system is stable; otherwise, it indicates that the grid-connected energy storage system is unstable.

[0205] In other embodiments, the short-circuit ratio is also dynamically evaluated, i.e.:

[0206]

[0207] In the formula, is the short - circuit ratio at time t, is the current vector injected by the grid - forming energy storage system into the power grid.

[0208] Among them, the larger the short - circuit ratio, the more stable the power grid is, and it can better stabilize the power fluctuations of energy storage or load. In one example, > 3.0 is the normal VSG mode, 1.5 < SCR ≤ 3.0 is the enhanced damping mode, and SCR ≤ 1.5 is the load - shedding / power - limiting mode.

[0209] In addition, the stability of the system can also be evaluated through the global performance index That is:

[0210]

[0211] In the formula, is the voltage recovery time, generally the time from the occurrence of the disturbance to the voltage recovering to ± 5% of the rated value. The smaller the global performance index is, the faster the system can recover stability with lower energy consumption

[0212] Based on the same inventive concept, the embodiment of the present application also provides a grid - forming energy storage collaborative control system for implementing the grid - forming energy storage collaborative control method involved above.

[0213] The implementation solutions provided by this system to solve problems are similar to those recorded in the above - mentioned method. Therefore, the specific limitations in one or more embodiments of the grid - forming energy storage collaborative control system provided below can be referred to the limitations of the grid - forming energy storage collaborative control method in the above text, and will not be elaborated here.

[0214] As [[ID=3,2]] Figure 3 [[ID=3,3]]shown, the embodiment of the present application also provides a grid - forming energy storage collaborative control system, including:

[0215] A dynamic admittance construction module 100, configured to construct a dynamic admittance matrix according to the hybrid topological parameters between the star - shaped main grid and the honeycomb unit in the power grid topology where the grid - forming energy storage system is located; wherein, the dynamic admittance matrix is used to characterize the impedance coupling relationship between the star - shaped main grid and the honeycomb unit at multiple spatio - temporal scales; [[ID=3,7]] [[ID=3,8]]

[0216] [[ID=3,9]]A fault flag determination module 200, configured to obtain the voltage measurement data of the common connection point of the grid - forming energy storage system, and determine the fault flag bit of the grid - forming energy storage system in combination with the dynamic admittance matrix; wherein, the fault flag bit is used to characterize the dual - fault characteristics of grid voltage drop and current over - limit; [[ID=4,]] [[ID=4,1]]

[0217] [[ID=4,2]]A control parameter optimization module 300, configured to adjust the fractional - order VSG control parameters of the grid - forming energy storage system based on the reinforcement learning algorithm in combination with the fault flag bit; [[ID=4,3]] [[ID=4,4]]

[0218] The state estimation module 400 is used to acquire the observed state data and fractional-order VSG control parameters of each energy storage unit in the grid-connected energy storage system, and to perform state estimation on the observed state data by combining the improved multi-rate Runge-Kutta method to obtain the state estimation value of each energy storage unit.

[0219] The power allocation execution module 500 is used to optimize the power allocation according to the state estimate of each energy storage unit based on the total output power of the grid-connected energy storage system in advance, generate power commands for each energy storage unit, and send the power commands of each energy storage unit to the corresponding energy storage units for execution.

[0220] In some embodiments, the hybrid topology parameters include the line impedance of the star-shaped main grid, the real-time virtual impedance of the grid-connected energy storage system, the filter inductance, and the grid angular frequency.

[0221] Dynamic admittance building module 100 is used for:

[0222] Obtain the topological connection structure between the star main grid and the honeycomb unit within the power grid topology where the grid-connected energy storage system is located, and identify the physical coupling relationship between the star main grid and the honeycomb unit;

[0223] Based on the real-time virtual impedance, filter inductance, and grid angular frequency of the grid-connected energy storage system, the time-varying admittance matrix of the honeycomb unit is constructed.

[0224] Determine the real admittance matrix of the star-shaped main network based on the line impedance of the star-shaped main network.

[0225] Based on the physical coupling relationship between the star-shaped main grid and the honeycomb unit, and combining the direct admittance from the star-shaped main grid to the honeycomb unit and the time-varying admittance matrix of the honeycomb unit, the coupling admittance matrix between the star-shaped main grid and the honeycomb unit is determined; where the direct admittance is the complex admittance of the physical branch connection between the star-shaped main grid and the honeycomb unit.

[0226] The time-varying admittance matrix of the honeycomb unit, the real admittance matrix of the star main grid, and the coupled admittance matrix are integrated in the form of a block matrix to synthesize a dynamic admittance matrix.

[0227] In some embodiments, the fault flag determination module 200 is configured to:

[0228] Obtain voltage measurement data at the common connection point of the grid-connected energy storage system;

[0229] Based on voltage measurement data and dynamic admittance matrix, calculate the injection current vector at the common connection point of the grid-connected energy storage system;

[0230] Based on voltage measurement data and injected current vector, determine whether the grid-connected energy storage system has experienced both voltage drop and current over-limit faults.

[0231] If it is determined that the grid-connected energy storage system has experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 1;

[0232] If it is determined that the grid-connected energy storage system has not experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 0.

[0233] In some embodiments, the fractional-order VSG control parameters include the fractional-order order;

[0234] The control parameter optimization module 300 is used for:

[0235] The state space and action space of the reinforcement learning algorithm are determined based on the fractional order and the order adjustment amount.

[0236] The reward function is determined based on the fault flag bit, combined with the voltage deviation and short-circuit ratio of the grid-connected energy storage system;

[0237] Based on the reward function, and combining the state space and action space, the fractional order is updated using the policy gradient method until the convergence condition is met. The latest fractional order is then output as the adjusted fractional VSG control parameter.

[0238] In some embodiments, the energy storage unit includes a lithium battery subsystem and a supercapacitor subsystem; the state estimate includes the state of charge estimate of the lithium battery subsystem and the voltage estimate of the supercapacitor subsystem.

[0239] State estimation module 400 is used for:

[0240] Acquire observational status data for the lithium battery subsystem and the supercapacitor subsystem; the observational status data includes the rate of change of state of charge of the lithium battery subsystem and the port voltage of the supercapacitor subsystem.

[0241] Based on the rate of change of state of charge of the lithium battery subsystem, a third-order Runge-Kutta explicit iterative scheme is constructed based on the implicit correction mechanism, and implicit correction coefficients and adaptive step size are introduced to predict the estimated state of charge of the lithium battery subsystem.

[0242] Based on the port voltage and fractional-order VSG control parameters of the supercapacitor subsystem, determine the fractional-order derivative term of the supercapacitor subsystem;

[0243] Based on the fractional derivative term of the supercapacitor subsystem, a fourth-order Runge-Kutta implicit iterative scheme is constructed using an explicit correction mechanism. Implicit correction coefficients and adaptive step size are introduced to predict the voltage estimate of the supercapacitor subsystem.

[0244] In some embodiments, the power allocation execution module 500 is configured to:

[0245] Based on the response characteristics of the supercapacitor subsystem and the lithium battery subsystem, the high-frequency power component of the total output power is allocated to the supercapacitor subsystem, and the low-frequency power component is allocated to the lithium battery subsystem; wherein, the high-frequency power component and the low-frequency power component are obtained by performing fractional differential filtering on the total output power.

[0246] The theoretical value of power distribution of the supercapacitor subsystem is determined based on the total output power and the response time constant of the supercapacitor subsystem.

[0247] Based on the voltage estimate of the supercapacitor subsystem, the safe power limit of the supercapacitor subsystem is determined, and the minimum value between the theoretical power allocation value and the safe power limit is taken as the power allocation value of the supercapacitor subsystem.

[0248] Based on the total output power and the theoretical power allocation value of the supercapacitor subsystem, the remaining power is determined as the theoretical power allocation value of the lithium battery subsystem.

[0249] Based on the theoretical power allocation value of the lithium battery subsystem and the estimated state of charge value of the lithium battery subsystem, the power allocation value of the lithium battery subsystem is determined.

[0250] Based on the power allocation values ​​of the supercapacitor subsystem and the lithium battery subsystem, power commands for the supercapacitor subsystem and the lithium battery subsystem are generated respectively.

[0251] The power commands for the supercapacitor subsystem and the lithium battery subsystem are sent to the supercapacitor subsystem and the lithium battery subsystem respectively for execution.

[0252] In some embodiments, the system further includes a stability evaluation module, used for:

[0253] The initial VSG virtual impedance of the grid-connected energy storage system is obtained, and the initial VSG virtual impedance is corrected by a fractional order to obtain the VSG virtual impedance after fractional order correction.

[0254] Based on the fractional-order corrected VSG virtual impedance and the equivalent capacitance of the supercapacitor subsystem, the impedance partial derivative of the supercapacitor subsystem is determined.

[0255] Based on the fractional-order corrected VSG virtual impedance, combined with the equivalent internal resistance and nominal capacity of the lithium battery subsystem, the impedance partial derivative of the lithium battery subsystem is determined.

[0256] The Jacobian matrix is ​​constructed using the impedance partial derivatives of the supercapacitor subsystem and the lithium battery subsystem.

[0257] Based on the Jacobian matrix, the maximum Lyapunov exponent is determined, and the stability of the grid-connected energy storage system is assessed based on the maximum Lyapunov exponent.

[0258] like Figure 4 As shown, this application embodiment provides an electronic device. The electronic device 10 includes a memory 20 and a processor 30. The memory 20 stores a computer program. When the computer program is executed by the processor 30, the processor 30 performs the following steps:

[0259] Based on the hybrid topology parameters between the star main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located, a dynamic admittance matrix is ​​constructed. The dynamic admittance matrix is ​​used to characterize the impedance coupling relationship between the star main grid and the honeycomb units at multiple spatiotemporal scales.

[0260] The voltage measurement data of the common coupling point of the grid-connected energy storage system is obtained, and the fault flag bit of the grid-connected energy storage system is determined by combining the dynamic admittance matrix; the fault flag bit is used to characterize the dual fault characteristics of grid voltage drop and current over-limit.

[0261] Based on reinforcement learning algorithms, the fractional-order VSG control parameters of the grid-connected energy storage system are adjusted in combination with fault flag bits.

[0262] The observed state data and fractional-order VSG control parameters of each energy storage unit in the grid-connected energy storage system are obtained. The state estimation of the observed state data is performed by combining the improved multi-rate Runge-Kutta method to obtain the state estimate value of each energy storage unit.

[0263] Based on the total output power of the grid-connected energy storage system obtained in advance, power allocation is optimized according to the state estimates of each energy storage unit, power commands for each energy storage unit are generated, and the power commands of each energy storage unit are sent to the corresponding energy storage units for execution.

[0264] In some embodiments, the hybrid topology parameters include the line impedance of the star-shaped main grid, the real-time virtual impedance of the grid-connected energy storage system, the filter inductance, and the grid angular frequency.

[0265] When the computer program is executed by the processor 30, the processor 30 also performs the following steps:

[0266] Obtain the topological connection structure between the star main grid and the honeycomb unit within the power grid topology where the grid-connected energy storage system is located, and identify the physical coupling relationship between the star main grid and the honeycomb unit;

[0267] Based on the real-time virtual impedance, filter inductance, and grid angular frequency of the grid-connected energy storage system, the time-varying admittance matrix of the honeycomb unit is constructed.

[0268] Determine the real admittance matrix of the star-shaped main network based on the line impedance of the star-shaped main network.

[0269] Based on the physical coupling relationship between the star-shaped main grid and the honeycomb unit, and combining the direct admittance from the star-shaped main grid to the honeycomb unit and the time-varying admittance matrix of the honeycomb unit, the coupling admittance matrix between the star-shaped main grid and the honeycomb unit is determined; where the direct admittance is the complex admittance of the physical branch connection between the star-shaped main grid and the honeycomb unit.

[0270] The time-varying admittance matrix of the honeycomb unit, the real admittance matrix of the star main grid, and the coupled admittance matrix are integrated in the form of a block matrix to synthesize a dynamic admittance matrix.

[0271] In some embodiments, when the computer program is executed by the processor 30, the processor 30 further performs the following steps:

[0272] Obtain voltage measurement data at the common connection point of the grid-connected energy storage system;

[0273] Based on voltage measurement data and dynamic admittance matrix, calculate the injection current vector at the common connection point of the grid-connected energy storage system;

[0274] Based on voltage measurement data and injected current vector, determine whether the grid-connected energy storage system has experienced both voltage drop and current over-limit faults.

[0275] If it is determined that the grid-connected energy storage system has experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 1;

[0276] If it is determined that the grid-connected energy storage system has not experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 0.

[0277] In some embodiments, the fractional-order VSG control parameters include the fractional-order order;

[0278] Based on reinforcement learning algorithms, and combined with fault flag bits, the fractional-order VSG control parameters of the grid-connected energy storage system are adjusted, including:

[0279] The state space and action space of the reinforcement learning algorithm are determined based on the fractional order and the order adjustment amount.

[0280] The reward function is determined based on the fault flag bit, combined with the voltage deviation and short-circuit ratio of the grid-connected energy storage system;

[0281] Based on the reward function, and combining the state space and action space, the fractional order is updated using the policy gradient method until the convergence condition is met. The latest fractional order is then output as the adjusted fractional VSG control parameter.

[0282] In some embodiments, the energy storage unit includes a lithium battery subsystem and a supercapacitor subsystem; the state estimate includes the state of charge estimate of the lithium battery subsystem and the voltage estimate of the supercapacitor subsystem.

[0283] When the computer program is executed by the processor 30, the processor 30 also performs the following steps:

[0284] Acquire observational status data for the lithium battery subsystem and the supercapacitor subsystem; the observational status data includes the rate of change of state of charge of the lithium battery subsystem and the port voltage of the supercapacitor subsystem.

[0285] Based on the rate of change of state of charge of the lithium battery subsystem, a third-order Runge-Kutta explicit iterative scheme is constructed based on the implicit correction mechanism, and implicit correction coefficients and adaptive step size are introduced to predict the estimated state of charge of the lithium battery subsystem.

[0286] Based on the port voltage and fractional-order VSG control parameters of the supercapacitor subsystem, determine the fractional-order derivative term of the supercapacitor subsystem;

[0287] Based on the fractional derivative term of the supercapacitor subsystem, a fourth-order Runge-Kutta implicit iterative scheme is constructed using an explicit correction mechanism. Implicit correction coefficients and adaptive step size are introduced to predict the voltage estimate of the supercapacitor subsystem.

[0288] In some embodiments, when the computer program is executed by the processor 30, the processor 30 further performs the following steps:

[0289] Based on the response characteristics of the supercapacitor subsystem and the lithium battery subsystem, the high-frequency power component of the total output power is allocated to the supercapacitor subsystem, and the low-frequency power component is allocated to the lithium battery subsystem; wherein, the high-frequency power component and the low-frequency power component are obtained by performing fractional differential filtering on the total output power.

[0290] The theoretical value of power distribution of the supercapacitor subsystem is determined based on the total output power and the response time constant of the supercapacitor subsystem.

[0291] Based on the voltage estimate of the supercapacitor subsystem, the safe power limit of the supercapacitor subsystem is determined, and the minimum value between the theoretical power allocation value and the safe power limit is taken as the power allocation value of the supercapacitor subsystem.

[0292] Based on the total output power and the theoretical power allocation value of the supercapacitor subsystem, the remaining power is determined as the theoretical power allocation value of the lithium battery subsystem.

[0293] Based on the theoretical power allocation value of the lithium battery subsystem and the estimated state of charge value of the lithium battery subsystem, the power allocation value of the lithium battery subsystem is determined.

[0294] Based on the power allocation values ​​of the supercapacitor subsystem and the lithium battery subsystem, power commands for the supercapacitor subsystem and the lithium battery subsystem are generated respectively.

[0295] The power commands for the supercapacitor subsystem and the lithium battery subsystem are sent to the supercapacitor subsystem and the lithium battery subsystem respectively for execution.

[0296] In some embodiments, when the computer program is executed by the processor 30, the processor 30 further performs the following steps:

[0297] The initial VSG virtual impedance of the grid-connected energy storage system is obtained, and the initial VSG virtual impedance is corrected by a fractional order to obtain the VSG virtual impedance after fractional order correction.

[0298] Based on the fractional-order corrected VSG virtual impedance and the equivalent capacitance of the supercapacitor subsystem, the impedance partial derivative of the supercapacitor subsystem is determined.

[0299] Based on the fractional-order corrected VSG virtual impedance, combined with the equivalent internal resistance and nominal capacity of the lithium battery subsystem, the impedance partial derivative of the lithium battery subsystem is determined.

[0300] The Jacobian matrix is ​​constructed using the impedance partial derivatives of the supercapacitor subsystem and the lithium battery subsystem.

[0301] Based on the Jacobian matrix, the maximum Lyapunov exponent is determined, and the stability of the grid-connected energy storage system is assessed based on the maximum Lyapunov exponent.

[0302] This application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed, it performs the following steps:

[0303] Based on the hybrid topology parameters between the star main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located, a dynamic admittance matrix is ​​constructed. The dynamic admittance matrix is ​​used to characterize the impedance coupling relationship between the star main grid and the honeycomb units at multiple spatiotemporal scales.

[0304] The voltage measurement data of the common coupling point of the grid-connected energy storage system is obtained, and the fault flag bit of the grid-connected energy storage system is determined by combining the dynamic admittance matrix; the fault flag bit is used to characterize the dual fault characteristics of grid voltage drop and current over-limit.

[0305] Based on reinforcement learning algorithms, the fractional-order VSG control parameters of the grid-connected energy storage system are adjusted in combination with fault flag bits.

[0306] The observed state data and fractional-order VSG control parameters of each energy storage unit in the grid-connected energy storage system are obtained. The state estimation of the observed state data is performed by combining the improved multi-rate Runge-Kutta method to obtain the state estimate value of each energy storage unit.

[0307] Based on the total output power of the grid-connected energy storage system obtained in advance, power allocation is optimized according to the state estimates of each energy storage unit, power commands for each energy storage unit are generated, and the power commands of each energy storage unit are sent to the corresponding energy storage units for execution.

[0308] In some embodiments, the hybrid topology parameters include the line impedance of the star-shaped main grid, the real-time virtual impedance of the grid-connected energy storage system, the filter inductance, and the grid angular frequency.

[0309] When a computer program is executed, it performs the following steps:

[0310] Obtain the topological connection structure between the star main grid and the honeycomb unit within the power grid topology where the grid-connected energy storage system is located, and identify the physical coupling relationship between the star main grid and the honeycomb unit;

[0311] Based on the real-time virtual impedance, filter inductance, and grid angular frequency of the grid-connected energy storage system, the time-varying admittance matrix of the honeycomb unit is constructed.

[0312] Determine the real admittance matrix of the star-shaped main network based on the line impedance of the star-shaped main network.

[0313] Based on the physical coupling relationship between the star-shaped main grid and the honeycomb unit, and combining the direct admittance from the star-shaped main grid to the honeycomb unit and the time-varying admittance matrix of the honeycomb unit, the coupling admittance matrix between the star-shaped main grid and the honeycomb unit is determined; where the direct admittance is the complex admittance of the physical branch connection between the star-shaped main grid and the honeycomb unit.

[0314] The time-varying admittance matrix of the honeycomb unit, the real admittance matrix of the star main grid, and the coupled admittance matrix are integrated in the form of a block matrix to synthesize a dynamic admittance matrix.

[0315] In some embodiments, when a computer program is executed, it performs the following steps:

[0316] Obtain voltage measurement data at the common connection point of the grid-connected energy storage system;

[0317] Based on voltage measurement data and dynamic admittance matrix, calculate the injection current vector at the common connection point of the grid-connected energy storage system;

[0318] Based on voltage measurement data and injected current vector, determine whether the grid-connected energy storage system has experienced both voltage drop and current over-limit faults.

[0319] If it is determined that the grid-connected energy storage system has experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 1;

[0320] If it is determined that the grid-connected energy storage system has not experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 0.

[0321] In some embodiments, the fractional-order VSG control parameters include the fractional-order order;

[0322] When a computer program is executed, it performs the following steps:

[0323] The state space and action space of the reinforcement learning algorithm are determined based on the fractional order and the order adjustment amount.

[0324] The reward function is determined based on the fault flag bit, combined with the voltage deviation and short-circuit ratio of the grid-connected energy storage system;

[0325] Based on the reward function, and combining the state space and action space, the fractional order is updated using the policy gradient method until the convergence condition is met. The latest fractional order is then output as the adjusted fractional VSG control parameter.

[0326] In some embodiments, the energy storage unit includes a lithium battery subsystem and a supercapacitor subsystem; the state estimate includes the state of charge estimate of the lithium battery subsystem and the voltage estimate of the supercapacitor subsystem.

[0327] When a computer program is executed, it performs the following steps:

[0328] Acquire observational status data for the lithium battery subsystem and the supercapacitor subsystem; the observational status data includes the rate of change of state of charge of the lithium battery subsystem and the port voltage of the supercapacitor subsystem.

[0329] Based on the rate of change of state of charge of the lithium battery subsystem, a third-order Runge-Kutta explicit iterative scheme is constructed based on the implicit correction mechanism, and implicit correction coefficients and adaptive step size are introduced to predict the estimated state of charge of the lithium battery subsystem.

[0330] Based on the port voltage and fractional-order VSG control parameters of the supercapacitor subsystem, determine the fractional-order derivative term of the supercapacitor subsystem;

[0331] Based on the fractional derivative term of the supercapacitor subsystem, a fourth-order Runge-Kutta implicit iterative scheme is constructed using an explicit correction mechanism. Implicit correction coefficients and adaptive step size are introduced to predict the voltage estimate of the supercapacitor subsystem.

[0332] In some embodiments, when a computer program is executed, it performs the following steps:

[0333] Based on the response characteristics of the supercapacitor subsystem and the lithium battery subsystem, the high-frequency power component of the total output power is allocated to the supercapacitor subsystem, and the low-frequency power component is allocated to the lithium battery subsystem; wherein, the high-frequency power component and the low-frequency power component are obtained by performing fractional differential filtering on the total output power.

[0334] The theoretical value of power distribution of the supercapacitor subsystem is determined based on the total output power and the response time constant of the supercapacitor subsystem.

[0335] Based on the voltage estimate of the supercapacitor subsystem, the safe power limit of the supercapacitor subsystem is determined, and the minimum value between the theoretical power allocation value and the safe power limit is taken as the power allocation value of the supercapacitor subsystem.

[0336] Based on the total output power and the theoretical power allocation value of the supercapacitor subsystem, the remaining power is determined as the theoretical power allocation value of the lithium battery subsystem.

[0337] Based on the theoretical power allocation value of the lithium battery subsystem and the estimated state of charge value of the lithium battery subsystem, the power allocation value of the lithium battery subsystem is determined.

[0338] Based on the power allocation values ​​of the supercapacitor subsystem and the lithium battery subsystem, power commands for the supercapacitor subsystem and the lithium battery subsystem are generated respectively.

[0339] The power commands for the supercapacitor subsystem and the lithium battery subsystem are sent to the supercapacitor subsystem and the lithium battery subsystem respectively for execution.

[0340] In some embodiments, when a computer program is executed, it performs the following steps:

[0341] The initial VSG virtual impedance of the grid-connected energy storage system is obtained, and the initial VSG virtual impedance is corrected by a fractional order to obtain the VSG virtual impedance after fractional order correction.

[0342] Based on the fractional-order corrected VSG virtual impedance and the equivalent capacitance of the supercapacitor subsystem, the impedance partial derivative of the supercapacitor subsystem is determined.

[0343] Based on the fractional-order corrected VSG virtual impedance, combined with the equivalent internal resistance and nominal capacity of the lithium battery subsystem, the impedance partial derivative of the lithium battery subsystem is determined.

[0344] The Jacobian matrix is ​​constructed using the impedance partial derivatives of the supercapacitor subsystem and the lithium battery subsystem.

[0345] Based on the Jacobian matrix, the maximum Lyapunov exponent is determined, and the stability of the grid-connected energy storage system is assessed based on the maximum Lyapunov exponent.

[0346] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, electronic devices, and computer storage media described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0347] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims and accompanying drawings of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such processes, methods, products or devices.

[0348] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0349] In the embodiments provided by this invention, it should be understood that the disclosed systems, electronic devices, computer storage media, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between devices or units through some interfaces, and may be electrical, mechanical, or other forms.

[0350] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0351] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0352] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for executing all or part of the steps of the methods described in the various embodiments of the present invention through a computer device (which may be a personal computer, a server, or a network device, etc.). The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0353] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for coordinated control of grid-connected energy storage, characterized in that, include: Based on the hybrid topology parameters between the star-shaped main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located, a dynamic admittance matrix is ​​constructed; wherein, the dynamic admittance matrix is ​​used to characterize the impedance coupling relationship between the star-shaped main grid and the honeycomb units under multiple spatiotemporal scales; The voltage measurement data of the common connection point of the grid-connected energy storage system is obtained, and the fault flag bit of the grid-connected energy storage system is determined by combining the dynamic admittance matrix; wherein, the fault flag bit is used to characterize the dual fault characteristics of grid voltage drop and current over-limit. Based on reinforcement learning algorithms, the fractional-order VSG control parameters of the grid-connected energy storage system are adjusted in conjunction with the fault flag bits. The observed state data of each energy storage unit in the grid-connected energy storage system and the fractional-order VSG control parameters are obtained. The state estimation of the observed state data is performed by combining the improved multi-rate Runge-Kutta method to obtain the state estimation value of each energy storage unit. Based on the total output power of the grid-connected energy storage system obtained in advance, power allocation optimization is performed according to the state estimation values ​​of each energy storage unit, power commands for each energy storage unit are generated, and the power commands for each energy storage unit are sent to the corresponding energy storage units for execution.

2. The grid-connected energy storage coordinated control method according to claim 1, characterized in that, The hybrid topology parameters include the line impedance of the star-shaped main grid, the real-time virtual impedance of the grid-connected energy storage system, the filter inductance, and the grid angular frequency. Based on the hybrid topology parameters between the star-shaped main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located, a dynamic admittance matrix is ​​constructed, including: Obtain the topological connection structure between the star main grid and the honeycomb unit within the power grid topology where the grid-connected energy storage system is located, and identify the physical coupling relationship between the star main grid and the honeycomb unit; Based on the real-time virtual impedance, filter inductance, and grid angular frequency of the grid-connected energy storage system, the time-varying admittance matrix of the honeycomb unit is constructed. Based on the line impedance of the star-shaped main network, determine the real admittance matrix of the star-shaped main network. Based on the physical coupling relationship between the star-shaped main grid and the honeycomb unit, and combining the direct admittance from the star-shaped main grid to the honeycomb unit and the time-varying admittance matrix of the honeycomb unit, the coupling admittance matrix between the star-shaped main grid and the honeycomb unit is determined; wherein, the direct admittance is the complex admittance of the physical branch connection between the star-shaped main grid and the honeycomb unit; The time-varying admittance matrix of the honeycomb unit, the real admittance matrix of the star-shaped main grid, and the coupling admittance matrix are integrated in the form of a block matrix to synthesize the dynamic admittance matrix.

3. The grid-connected energy storage coordinated control method according to claim 1, characterized in that, The step of obtaining voltage measurement data at the common coupling point of the grid-connected energy storage system and determining the fault flag bit of the grid-connected energy storage system in conjunction with the dynamic admittance matrix includes: Obtain voltage measurement data at the common connection point of the grid-connected energy storage system; Based on the voltage measurement data and the dynamic admittance matrix, calculate the injection current vector at the common connection point of the grid-connected energy storage system; Based on the voltage measurement data and the injected current vector, it is determined whether the grid-connected energy storage system has experienced both voltage drop and current over-limit faults. If it is determined that the grid-connected energy storage system has experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 1; If it is determined that the grid-connected energy storage system has not experienced both voltage drop and current over-limit faults, then the fault flag bit is set to 0.

4. The grid-connected energy storage coordinated control method according to claim 1, characterized in that, The fractional-order VSG control parameters include the fractional-order order; The step of adjusting the fractional-order VSG control parameters of the grid-connected energy storage system based on the reinforcement learning algorithm and in conjunction with the fault flag bits includes: The state space and action space of the reinforcement learning algorithm are determined based on the fractional order and the order adjustment amount. Based on the fault flag, and in conjunction with the voltage deviation and short-circuit ratio of the grid-connected energy storage system, the reward function is determined; Based on the reward function, and combining the state space and the action space, the fractional order is updated using the policy gradient method until the convergence condition is met. The latest fractional order is then output as the adjusted fractional VSG control parameter.

5. The grid-connected energy storage coordinated control method according to claim 1 or 4, characterized in that, The energy storage unit includes a lithium battery subsystem and a supercapacitor subsystem; the state estimate includes the state of charge estimate of the lithium battery subsystem and the voltage estimate of the supercapacitor subsystem; The process of acquiring the observed state data of each energy storage unit in the grid-connected energy storage system and the fractional-order VSG control parameters, and then using the improved multi-rate Runge-Kutta method to perform state estimation on the observed state data to obtain the state estimate value of each energy storage unit includes: Acquire observational status data of the lithium battery subsystem and the supercapacitor subsystem; wherein, the observational status data includes the state-of-charge rate of the lithium battery subsystem and the port voltage of the supercapacitor subsystem; Based on the rate of change of state of charge of the lithium battery subsystem, a third-order Runge-Kutta explicit iterative scheme is constructed based on the implicit correction mechanism, and implicit correction coefficients and adaptive step size are introduced to predict the estimated state of charge of the lithium battery subsystem. The fractional derivative term of the supercapacitor subsystem is determined based on the port voltage of the supercapacitor subsystem and the fractional-order VSG control parameters. Based on the fractional derivative terms of the supercapacitor subsystem, a fourth-order Runge-Kutta implicit iterative scheme is constructed using an explicit correction mechanism. Implicit correction coefficients and adaptive step sizes are introduced to predict the voltage estimate of the supercapacitor subsystem.

6. The grid-connected energy storage coordinated control method according to claim 5, characterized in that, The step of optimizing power allocation based on the pre-acquired total output power of the grid-connected energy storage system and the state estimates of each energy storage unit, generating power commands for each energy storage unit, and issuing the power commands to the corresponding energy storage units for execution includes: Based on the response characteristics of the supercapacitor subsystem and the lithium battery subsystem, the high-frequency power component of the total output power is allocated to the supercapacitor subsystem, and the low-frequency power component is allocated to the lithium battery subsystem; wherein, the high-frequency power component and the low-frequency power component are obtained by performing fractional-order differential filtering on the total output power; Based on the total output power and the response time constant of the supercapacitor subsystem, the theoretical value of the power distribution of the supercapacitor subsystem is determined; Based on the voltage estimate of the supercapacitor subsystem, the safe power limit of the supercapacitor subsystem is determined, and the minimum value between the theoretical power allocation value and the safe power limit is taken as the power allocation value of the supercapacitor subsystem. Based on the total output power and the theoretical power allocation value of the supercapacitor subsystem, the remaining power is determined as the theoretical power allocation value of the lithium battery subsystem; Based on the theoretical power allocation value of the lithium battery subsystem and the estimated state of charge value of the lithium battery subsystem, the power allocation value of the lithium battery subsystem is determined. Based on the power allocation values ​​of the supercapacitor subsystem and the lithium battery subsystem, power commands for the supercapacitor subsystem and the lithium battery subsystem are generated respectively. The power commands of the supercapacitor subsystem and the lithium battery subsystem are respectively sent to the supercapacitor subsystem and the lithium battery subsystem for execution.

7. The grid-connected energy storage coordinated control method according to claim 5, characterized in that, Also includes: The initial VSG virtual impedance of the grid-connected energy storage system is obtained, and the initial VSG virtual impedance is corrected by a fractional order to obtain the VSG virtual impedance after fractional order correction. Based on the fractionally corrected VSG virtual impedance and the equivalent capacitance of the supercapacitor subsystem, the impedance partial derivative of the supercapacitor subsystem is determined. Based on the fractionally corrected VSG virtual impedance, combined with the equivalent internal resistance and nominal capacity of the lithium battery subsystem, the impedance bias of the lithium battery subsystem is determined. The Jacobian matrix is ​​constructed using the impedance partial derivatives of the supercapacitor subsystem and the lithium battery subsystem. Based on the Jacobian matrix, the maximum Lyapunov exponent is determined, and the stability of the grid-connected energy storage system is evaluated based on the maximum Lyapunov exponent.

8. A grid-connected energy storage collaborative control system, characterized in that, include: The dynamic admittance construction module is used to construct a dynamic admittance matrix based on the mixed topology parameters between the star main grid and the honeycomb units within the power grid topology where the grid-connected energy storage system is located; wherein, the dynamic admittance matrix is ​​used to characterize the impedance coupling relationship between the star main grid and the honeycomb units at multiple spatiotemporal scales; The fault flag determination module is used to acquire voltage measurement data of the common connection point of the grid-connected energy storage system, and determine the fault flag bit of the grid-connected energy storage system in combination with the dynamic admittance matrix; wherein, the fault flag bit is used to characterize the dual fault characteristics of grid voltage drop and current over-limit. The control parameter optimization module is used to adjust the fractional-order VSG control parameters of the grid-connected energy storage system based on a reinforcement learning algorithm and in conjunction with the fault flag bit. The state estimation module is used to acquire the observed state data of each energy storage unit of the grid-connected energy storage system and the fractional-order VSG control parameters, and to perform state estimation on the observed state data by combining the improved multi-rate Runge-Kutta method to obtain the state estimation value of each energy storage unit. The power allocation execution module is used to optimize the power allocation according to the state estimate of each energy storage unit based on the total output power of the grid-connected energy storage system obtained in advance, generate power instructions for each energy storage unit, and send the power instructions of each energy storage unit to the corresponding energy storage unit for execution.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor performs the steps of the grid-based energy storage coordinated control method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, it implements the steps of the grid-based energy storage coordinated control method as described in any one of claims 1-7.

Images