Parameter optimization method for network-structured VRFB energy storage system

Abstract

The application provides a parameter optimization method of a network-constructed VRFB energy storage system, relates to the technical field of energy storage system optimization, and comprises the following steps: based on the physical architecture and electrical characteristics of the VRFB energy storage system, a multi-dimensional extended state space model containing structure parameters and dynamic operation states is established; the eigenvalues of a state matrix A are solved by substituting engineering parameters, and it is judged whether there are eigenvalues leading to system divergence, resonance or insufficient robustness; an optimization target is constructed based on the real part, the imaginary part and the damping ratio of the eigenvalues, and an intelligent algorithm is used to solve the optimal parameter combination. The application constructs an extended state space model, accurately depicts the multi-physical field coupling behavior of the system, provides a reliable foundation for parameter optimization, solves the parameter coupling problem through a multi-objective optimization function and a genetic algorithm, improves the comprehensive performance of the system, verifies the system in multiple working conditions with the aid of a simulation and an online correction mechanism, ensures that the system maintains an optimal state in a complex environment, and provides protection for the large-scale application of the system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage system optimization, and particularly relates to a parameter optimization method for a grid-forming VRFB energy storage system. BACKGROUND

[0002] With the transformation of global energy structure to clean, large-scale grid-connection of renewable energy (such as wind energy and solar energy) has become the core trend of the development of power systems. However, the gradual withdrawal of traditional synchronous generators leads to insufficient inertia support of the power grid, presenting the characteristics of "low inertia and low damping", which seriously threatens the frequency stability of the system. The grid-forming (GFM) energy storage system, through the virtual synchronous generator (VSG) control technology, simulates the frequency and phase modulation capability of the synchronous machine, and becomes a key technical path to improve the stability of the power grid. The all-vanadium redox flow battery (VRFB) has become an ideal choice for grid-forming energy storage systems due to its long cycle life, high safety, and power and capacity decoupling advantages.

[0003] In the prior art, for the parameter optimization of the grid-forming VRFB energy storage system, a systematic solution has not been formed to integrate the physical characteristics of the energy storage system, the control logic and the dynamic characteristics of the power grid. The limitations of traditional technologies make the system prone to problems such as insufficient stability margin and response speed lag in actual operation, which seriously restricts its large-scale application. Therefore, it is necessary to provide a parameter optimization method for a grid-forming VRFB energy storage system to solve the above technical problems. SUMMARY

[0004] The present application aims to provide a parameter optimization method for a grid-forming VRFB energy storage system, which realizes the global optimal configuration of the parameters of the grid-forming VRFB energy storage system through multi-objective optimization and multi-working condition closed-loop verification, and ensures the stable operation of the system under the conditions of load mutation and power grid failure.

[0005] To solve the above technical problems, the present application provides a parameter optimization method for a grid-forming VRFB energy storage system, which comprises the following steps:

[0006] S1, constructing an extended state space model:

[0007] Based on the physical architecture and electrical characteristics of the VRFB energy storage system, a multi-dimensional extended state space model containing structural parameters and dynamic operating states is established;

[0008] S2, eigenvalue calculation and stability criterion:

[0009] The eigenvalues of the state matrix are solved by substituting the engineering parameters, and it is determined whether there are eigenvalues that cause the system to diverge, resonate or lack robustness;

[0010] S3, multi-objective optimization function construction:

[0011] An optimization objective is constructed based on the real part, the imaginary part and the damping ratio of the eigenvalue, and an intelligent algorithm is used to solve the optimal parameter combination.

[0012] S4, optimization result verification and correction:

[0013] The parameter effectiveness is verified through multi-condition closed-loop simulation, and if the index is not met, the parameter correction is triggered.

[0014] As preferred, a multi-dimensional extended state space model including structural parameters and dynamic operating states is established:

[0015] The structural parameters and operating state parameters of the VRFB energy storage system are obtained and converted into a state vector; wherein the state vector includes the inverter output angular frequency w, the grid-connected point active power P, the grid-connected point reactive power Q, the inverter output current idq in the dq coordinate system, the voltage vdq output by the filter in the dq coordinate system, the voltage loop differential operators x1, x2, the current loop differential operators x3, x4, and the DC voltage Udc;

[0016] The state equation and the output equation are established; wherein A is a state matrix including the battery dynamic, the inverter control logic and the grid coupling term, B is an input matrix, C is an output matrix, D is a feedforward matrix, u is an input vector, and y is an output vector;

[0017] The state vector x and the input vector u are substituted into the state equation and the output equation to obtain the extended state space model.

[0018] As preferred, the state matrix eigenvalue is solved by substituting the engineering parameters, to determine whether there is an eigenvalue that causes the system to diverge, resonate or lack robustness, specifically as follows:

[0019] The engineering parameters are obtained; the engineering parameters include VRFB battery physical parameters, inverter control parameters, filter parameters, and thermodynamic parameters; the VRFB battery physical parameters include single battery capacity, internal resistance, electrolyte temperature T, Faraday constant F, battery state of charge SOC, gas constant R, number of battery plates in the battery, number of battery series, number of battery parallel, standard electrode potential of the battery E; the inverter control parameters include virtual moment of inertia G, damping coefficient D, and virtual impedance control parameters; the filter parameters include inductance value L1 on the inverter side, inductance value L2 on the grid side, filter capacitance value Cf, voltage and current static values Vdq0 in the dq coordinate output by the filter, grid angular frequency W, and filter capacitance value on the DC side; the thermodynamic parameters include specific heat capacity of electrolyte cp and heat dissipation coefficient h;

[0020] The engineering parameters are substituted into the state matrix A to solve the characteristic equation to obtain the eigenvalue of the state matrix, wherein for determining system stability, imaginary part for reflecting the oscillation characteristics;

[0021] According to the real part and the imaginary part of the eigenvalue, it is judged whether the system has divergence, resonance or insufficient robustness, specifically:

[0022] If the real part Re of the eigenvalue is greater than 0, The system is unstable and may lead to system divergence.

[0023] Obtain the current continuous set number of eigenvalues, calculate the difference between the real part of the set number of eigenvalues and 0 to obtain the real part difference; Calculate the standard deviation of all real part differences to obtain the real part trend value; Weighted calculation of the current real part difference and the real part trend value to obtain the real part evaluation value; If the real part evaluation value is less than the set proximity threshold, it means that the real part is close to 0.

[0024] If there is an eigenvalue with an imaginary part Im ≠0 and the real part is close to 0, it means that it may cause system resonance; If the distribution of the eigenvalue is too dispersed or close to the imaginary axis, it means that the robustness of the system may be insufficient.

[0025] As a preferred, based on the real part, imaginary part and damping ratio of the eigenvalue, an optimization target is constructed, and an intelligent algorithm is used to solve the optimal parameter combination, specifically as follows:

[0026] When the intelligent algorithm uses genetic algorithm for optimization, the specific optimization steps are as follows:

[0027] S31, set the algorithm parameters of the genetic algorithm, including population size N, iteration number T1, crossover probability pc, mutation probability pm;

[0028] S32, encode the parameter combination to be optimized using binary, identify the number of parameters to be optimized as m, and represent each parameter with k binary numbers, then the encoding length of each individual is L=m×k bits;

[0029] S33, randomly generate N individuals to form an initial population; Each individual corresponds to a set of parameter combinations, and the specific parameter values can be obtained by decoding;

[0030] S34, for each individual in the population, substitute the corresponding parameter combination into the state matrix, solve the characteristic equation to obtain the eigenvalue, and then calculate its fitness value according to the optimization objective function, the calculation formula is:

[0031] + + wherein w1, w2, w3 are weight coefficients, and w1+w2+w3=1, n is the number of eigenvalues of the state matrix, is the real part of the i-th eigenvalue, and the sum of squares of the real parts reflects the stability of the system, is the damping ratio corresponding to the i-th eigenvalue,

[0032] S35, selecting an individual from the parent generation for crossover operation according to the crossover probability pc;

[0033] S36, performing mutation operation on the newly generated offspring individual according to the mutation probability pm

[0034] S37, determining whether the termination condition is met, the termination condition including reaching the number of iterations and the improvement amplitude value being less than the improvement threshold value; if the termination condition is met, the algorithm stops, and outputs the parameter combination corresponding to the individual with the minimum fitness value in the current population as the optimal parameter combination; otherwise, returning to step S34 to continue iteration.

[0035] As preferred, the parameter effectiveness is verified by multi-working condition closed-loop simulation, and if the index is not met, the parameter correction is triggered, specifically as follows:

[0036] Working condition setting and simulation execution:

[0037] Using the optimized parameters for multi-working condition closed-loop simulation, setting the working condition, including load mutation and power grid fault;

[0038] Substituting the optimized parameters into the network-type VRFB energy storage system mathematical model including the VRFB stack model, the inverter model, the controller model and the power grid model to perform closed-loop simulation;

[0039] Performance index monitoring:

[0040] In the simulation process, the performance indexes of the system are monitored in real time, and the performance indexes include voltage stability, frequency stability and power tracking accuracy; for each performance index, a corresponding threshold value is set, including a direct current voltage deviation threshold value, a frequency deviation threshold value and a power tracking error threshold value;

[0041] If any index is greater than its corresponding threshold value, it is determined that the index does not meet the requirement:

[0042] When the index does not meet the requirement, the weight coefficients of the optimization objective function or the parameters of the intelligent algorithm are adjusted accordingly:

[0043] When the voltage stability or the frequency stability is greater than its corresponding threshold value, the weight of the real part of the eigenvalue strongly related to the system stability in the target optimization function is increased according to the set rule, and the crossover probability of the genetic algorithm is reduced;

[0044] ​If the power tracking accuracy is greater than its corresponding threshold value, the imaginary part weight of the characteristic value related to the system dynamic characteristic in the target function is increased according to the set rule, and the mutation probability is increased;

[0045] Using the adjusted weight coefficient and intelligent algorithm parameter, the genetic algorithm is re-run for optimization and solution to obtain a new parameter combination; then, multi-working condition closed loop simulation verification is performed again, and the above steps are repeated until all performance indexes meet the requirements.

[0046] As preferred, in the optimization result verification and correction step, an online monitoring and real-time correction mechanism is further included, which is applied in the monitoring center:

[0047] In the actual operation process of the system, the operating parameters of the real-time network type VRFB energy storage system are constructed by sensors;

[0048] After the system is optimized, the obtained operating parameters are stored in the database of the monitoring center;

[0049] The operating parameters are compared with the optimized operating parameters to calculate the deviation value; a time region in a set operating monitoring time length before the current time is recorded as an operating monitoring time zone; the deviation value at any collection time in the operating monitoring time zone is identified, and the deviation trend value and the deviation average value are calculated by using the standard deviation and average formula; the deviation evaluation value is obtained by weighted calculation of the deviation value, the deviation trend value and the deviation average value at the current time;

[0050] A corresponding threshold value is set for the deviation evaluation value of each operating parameter, and if the deviation evaluation value of the operating parameter is greater than the corresponding threshold value, it indicates that the system operating state deviates from the optimal state, and the real-time correction mechanism needs to be triggered and an adjustment instruction is generated;

[0051] The real-time correction mechanism adopts a PID control algorithm:

[0052] The operating parameter corresponding to the triggered real-time correction mechanism is recorded as u, and the current deviation value is The output of the PID algorithm is the parameter adjustment amount The calculation formula of the PID control algorithm is Wherein Kp, Ki and Kd are the proportional coefficient, integral coefficient and differential coefficient, respectively;

[0053] The adjustment instruction and the parameter adjustment amount are sent to the controller through the communication interface; the operating parameter is adjusted by the controller.

[0054] As preferred, the operating parameters of the real-time network type VRFB energy storage system are constructed by sensors, and the specific settings of the sensors are as follows:

[0055] The sensors are arranged at key positions of the grid-connected VRFB energy storage system:

[0056] The current-voltage sensor is arranged in the VRFB stack to collect voltage and current data during the charging and discharging process of the stack, and the state of charge of the electrolyte is calculated in real time by using the ampere-hour integration method and the open circuit voltage method.

[0057] The electrolyte temperature sensors are distributed in the stack and the electrolyte storage container to monitor the temperature change of the electrolyte.

[0058] The voltage sensors are installed at the output end of the stack, the input and output ends of the inverter and the grid connection point to measure the voltage value.

[0059] The current sensors are arranged on the stack and the inverter to monitor the charging and discharging current of the stack and the output current of the inverter.

[0060] The operating parameters collected by the sensors are processed by the data acquisition module to convert the analog signals into digital signals, and are transmitted to the monitoring center through wired or wireless communication.

[0061] The application provides a device comprising a memory and a processor, wherein the memory stores a computer program of the above-mentioned parameter optimization method, and the processor executes the program to realize the parameter optimization of the grid-connected VRFB energy storage system.

[0062] Compared with the related art, the parameter optimization method of the grid-connected VRFB energy storage system has the following beneficial effects:

[0063] 1. The application constructs an extended state space model containing stack structure parameters, control parameters and operating states, integrates electrochemical dynamics and grid coupling characteristics, breaks through the limitations of traditional simplified modeling, accurately describes the multi-physical field coupling behavior of the system, provides a reliable theoretical basis for parameter optimization, and avoids stability misjudgment caused by model distortion.

[0064] 2. The application constructs a multi-objective optimization function based on the real part, imaginary part and damping ratio of the eigenvalue, and realizes the global collaborative optimization of the stack structure and control parameters by combining the genetic algorithm, effectively solves the problems of divergence, resonance and insufficient robustness caused by strong coupling of parameters, considers the system stability, dynamic response and multi-working-condition adaptability, and improves the comprehensive performance.

[0065] 3. The application verifies the effectiveness of the parameters through multi-working-condition closed-loop simulation of load mutation and grid fault, combines online monitoring and PID dynamic correction mechanism, compares the deviation of operating parameters in real time and triggers adjustment, ensures that the system maintains the optimal state in the complex grid environment, breaks through the bottleneck of the existing method "offline optimization-online failure", and provides reliability guarantee for the large-scale application of the grid-connected VRFB energy storage system.

[0066] In summary, the application constructs an extended state space model, accurately describes the system multi-physical field coupling behavior, and provides a reliable basis for parameter optimization; through the multi-objective optimization function and genetic algorithm, the parameter coupling problem is solved, and the system comprehensive performance is improved; with the help of multi-working condition simulation verification and online correction mechanism, the system is ensured to maintain the optimal state in complex environment, and the scale application is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 A flowchart of a parameter optimization method of a network-constructed VRFB energy storage system provided by the application. DETAILED DESCRIPTION

[0068] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0069] The terms used in the present disclosure are merely for the purpose of describing specific embodiments and are not intended to limit the present disclosure. The singular forms "a", "an", and "the" used in the present disclosure and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0070] It should be understood that although the terms first, second, third, etc. may be used in the present disclosure to describe various information, these information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of the present disclosure, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as the first information. Depending on the context, the word "if" as used herein can be interpreted as "when" or "upon" or "in response to determining".

[0071] Please refer to Figure 1 A parameter optimization method of a network-constructed VRFB energy storage system, comprising the following steps:

[0072] S1, constructing an extended state space model:

[0073] Based on the physical architecture and electrical characteristics of the VRFB energy storage system, a multi-dimensional extended state space model containing structural parameters and dynamic operating states is established;

[0074] S2, eigenvalue calculation and stability criterion:

[0075] Substitute the engineering parameters to solve the eigenvalues of the state matrix A, and judge whether there is an eigenvalue that leads to system divergence, resonance or insufficient robustness;

[0076] S3, multi-objective optimization function construction:

[0077] Based on the real part, imaginary part and damping ratio of the eigenvalue, an optimization target is constructed, and an intelligent algorithm is used to solve the optimal parameter combination;

[0078] S4, optimization result verification and correction:

[0079] The parameter effectiveness is verified through multi-working condition closed loop simulation, and if the index is not met, the parameter correction is triggered.

[0080] In the present application, a multi-dimensional extended state space model containing structural parameters and dynamic operating state is established:

[0081] The structural parameters and operating state parameters of the VRFB energy storage system are obtained and converted into a state vector; wherein the state vector includes the inverter output angular frequency w, the grid-connected point active power P, the grid-connected point reactive power Q, the inverter output current idq in the dq coordinate system, the voltage vdq output through the filter in the dq coordinate system, the voltage ring differential operator x1, x2, the current loop differential operator x3, x4, and the DC voltage Udc; the state equation is established And the output equation ; wherein A is a state matrix containing the battery dynamic, inverter control logic and grid coupling term, B is an input matrix, C is an output matrix, D is a feedforward matrix, u is an input vector, and y is an output vector;

[0082] Substitute the state vector x and the input vector u into the state equation and the output equation to obtain the extended state space model.

[0083] It should be noted that the structural parameters (such as the number of series and parallel batteries) and operating state parameters (such as state of charge, temperature, etc.) of the VRFB energy storage system are collected, and these parameters are combined into a state vector for subsequent digital processing; the state equation and the output equation are established; the state matrix A covers the battery dynamic, inverter control logic and grid coupling term, which reflects the interaction of each part of the system; the input matrix B, the output matrix C and the feedforward matrix D are used to describe the input, the output and the direct influence of the input on the output respectively; the state vector x and the input vector u are substituted into the above equation, thereby obtaining an extended state space model that can fully reflect the characteristics of the system, providing a basis for subsequent parameter optimization and the like.

[0084] In the present application, substitute the engineering parameters to solve the eigenvalues of the state matrix, judge whether there is an eigenvalue that leads to system divergence, resonance or insufficient robustness, as follows:

[0085] Obtaining engineering parameters; the engineering parameters include VRFB stack physical parameters, inverter control parameters, filter parameters, thermodynamic parameters; the VRFB stack physical parameters include single stack capacity, internal resistance, electrolyte temperature T, Faraday constant F, battery state of charge SOC, gas constant R, the number of battery plates in the stack, the number of stack series, the number of stack parallel, battery standard electrode electromotive force E; the inverter control parameters include virtual rotational inertia G, damping coefficient D, virtual impedance control parameters (virtual resistance value, virtual inductance value); the filter parameters include inductance value L1 on the inverter side, inductance value L2 on the grid side, filter capacitance value Cf, static voltage and current value Vdq0 in dq coordinates of filter output, grid angular frequency W, filter capacitance value on the DC side; the thermodynamic parameters include electrolyte specific heat capacity cp, heat dissipation coefficient h;

[0086] Substitute the engineering parameters into the state matrix A, solve the characteristic equation , to obtain the eigenvalue of the state matrix , wherein is used to determine the system stability, and the imaginary part is used to reflect the oscillation characteristics;

[0087] According to the real part and the imaginary part of the eigenvalue, it is judged whether the system has divergence, resonance or insufficient robustness, specifically:

[0088] If the real part Re( ) of the eigenvalue is greater than 0, the system is unstable, which may lead to system divergence;

[0089] Obtaining a certain number of continuous eigenvalues, calculating the real part difference between the real part of the certain number of eigenvalues and 0 to obtain the real part difference; calculating the standard deviation of all real part differences to obtain the real part trend value; calculating the real part evaluation value by weighting the current real part difference and the real part trend value; if the real part evaluation value is less than the set proximity threshold, it indicates that the real part is close to 0;

[0090] If there is an eigenvalue with a non-zero imaginary part Im( ) and the real part is close to 0, it indicates that it may cause system resonance; if the distribution of the eigenvalues is too dispersed (i.e. the standard deviation of the continuous multiple generation fitness values is greater than the preset trend threshold) or close to the imaginary axis (i.e. the absolute value of the real part of the eigenvalue is less than the set near-imaginary threshold, or the damping ratio is less than the damping threshold), it indicates that the robustness of the system may be insufficient.

[0091] It should be noted that the method realizes the accurate positioning of the potential risks of the network type VRFB energy storage system through the multi-dimensional analysis of the real part, the imaginary part and the distribution characteristics of the eigenvalue: the system instability caused by the direct correlation of the divergence risk parameter; the potential coupling problem of the control parameter and the grid frequency revealed by the resonance risk; the adaptability defect of the parameter combination to the working condition change is pointed out by the insufficient robustness; the three together constitute a complete evaluation system of system stability, which provides a clear correction direction for subsequent multi-objective optimization.

[0092] In the present application, the optimization target is constructed based on the real part, the imaginary part and the damping ratio of the eigenvalue, and the intelligent algorithm is used to solve the optimal parameter combination, as follows:

[0093] When the intelligent algorithm uses genetic algorithm for optimization, the specific optimization steps are as follows:

[0094] S31, set the algorithm parameters of genetic algorithm, including population size N, iteration number T1, crossover probability pc, mutation probability pm;

[0095] S32, encode the parameter combination to be optimized using binary, identify the number of parameters to be optimized as m, and represent each parameter with k binary numbers, then the encoding length of each individual is L=m×k bits;

[0096] S33, randomly generate N individuals to form an initial population; each individual corresponds to a set of parameter combinations, and the specific parameter values can be obtained by decoding;

[0097] S34, for each individual in the population, substitute the corresponding parameter combination into the state matrix, solve the characteristic equation to obtain the eigenvalue, and then calculate the fitness value according to the optimization objective function, the calculation formula of the optimization objective function is:

[0098] + + , wherein w1, w2, w3 are weight coefficients, and w1+w2+w3=1, n is the number of eigenvalues of the state matrix, is the real part of the i-th eigenvalue, and the square sum of the real part reflects the stability of the system, wherein the real part is less than 0, indicating that the system is stable, and the larger the absolute value of the real part, the better the stability of the system, is the damping ratio corresponding to the i-th eigenvalue, , the damping ratio is used to reflect the damping characteristics of the system, the closer the damping ratio is to 1, the smaller the oscillation of the system, and the more stable the response; the smaller the fitness value, the better the parameter combination corresponding to the individual;

[0099] S35, select individuals from the parent generation for crossover operation according to the crossover probability pc;

[0100] S36, according to the mutation probability pm, the new generated offspring individual is subjected to a mutation operation

[0101] S37, it is judged whether the termination condition is met, the termination condition includes reaching the iteration number and the improvement amplitude value being less than the improvement threshold value;If the termination condition is met, the algorithm stops, and the parameter combination corresponding to the individual with the minimum fitness value in the current population is output as the optimal parameter combination;Otherwise, return to step S34 to continue iteration.

[0102] The improvement amplitude value is obtained as follows:

[0103] The fitness values of the optimal solutions of the continuous generations are obtained, and the fitness values of the continuous optimal solutions are calculated by using a standard deviation formula to obtain the improvement amplitude value.

[0104] It should be noted that, by the optimization method based on the genetic algorithm, the optimal parameter combination of the grid-connected VRFB energy storage system can be effectively found by comprehensively considering the characteristic value related indexes, and the performance and stability of the system are improved.

[0105] In the present application, the parameter effectiveness is verified by multi-working condition closed loop simulation, and if the indexes are not met, the parameter correction is triggered, as follows:

[0106] Working condition setting and simulation execution:

[0107] Multi-working condition closed loop simulation is performed using the optimized parameters, and the working condition is set, including load mutation and power grid fault;

[0108] The optimized parameters are substituted into the grid-connected VRFB energy storage system mathematical model including the VRFB stack model, the inverter model, the controller model and the power grid model for closed loop simulation;

[0109] Performance index monitoring:

[0110] In the simulation process, the performance indexes of the system are monitored in real time, and the performance indexes include voltage stability, frequency stability and power tracking accuracy;For each performance index, a corresponding threshold value is set, including a DC voltage deviation threshold value, a frequency deviation threshold value and a power tracking error threshold value;

[0111] If any index is greater than its corresponding threshold value, it is determined that the index does not meet the requirements:

[0112] When the index does not meet the requirements, the weight coefficient of the optimization objective function or the parameters of the intelligent algorithm are adjusted accordingly:

[0113] When the voltage stability or the frequency stability is greater than its corresponding threshold value, the weight of the real part of the characteristic value related to the system stability in the target optimization function is increased according to the set rule, and the crossover probability of the genetic algorithm is reduced;

[0114] If the power tracking accuracy is greater than its corresponding threshold value, the imaginary part weight of the characteristic value related to the system dynamic characteristic in the target function is increased according to the set rule, and the mutation probability is increased;

[0115] Using the adjusted weight coefficient and intelligent algorithm parameter, the genetic algorithm is re-run for optimization and solution to obtain a new parameter combination; then multi-working condition closed loop simulation verification is carried out again, and the above steps are repeated until all performance indexes meet the requirements.

[0116] It should be noted that through multi-working condition closed loop simulation and dynamic correction, the present application realizes full-process coverage of parameter optimization from "theoretical modeling" to "engineering verification", ensures the robustness of the system under multi-dimensional risk scenarios, and lays a key technical foundation for the large-scale deployment of networked energy storage technology.

[0117] In the present application, in the optimization result verification and correction step, an online monitoring and real-time correction mechanism is also included, which is applied in the monitoring center:

[0118] In the actual operation process of the system, the running parameters of the networked VRFB energy storage system are monitored in real time through sensors;

[0119] After the system is optimized, the obtained running parameters are stored in the database of the monitoring center; after the monitoring center receives the real-time running parameters, the real-time running parameters are synchronized with the optimized running parameters stored in the database;

[0120] The running parameters and the optimized running parameters are compared to calculate the deviation value; a time region within a set running monitoring time length before the current time is recorded as a running monitoring time zone; the deviation value at any collection time in the running monitoring time zone is identified, and the deviation trend value and the deviation average value are calculated by using the standard deviation and average formula; the deviation value, the deviation trend value and the deviation average value at the current time are weighted to obtain a deviation evaluation value;

[0121] A corresponding threshold value is set for the deviation evaluation value of each running parameter, and if the deviation evaluation value of the running parameter is greater than the corresponding threshold value, it indicates that the system running state deviates from the optimal state, and the real-time correction mechanism needs to be triggered and an adjustment instruction is generated;

[0122] The real-time correction mechanism adopts a PID control algorithm:

[0123] The running parameter corresponding to the triggered real-time correction mechanism is recorded as u, and the current deviation value The output of the PID algorithm is the parameter adjustment amount The calculation formula of the PID control algorithm is , wherein Kp, Ki and Kd are proportional coefficient, integral coefficient and differential coefficient, respectively.

[0124] The adjustment instruction and the parameter adjustment amount are sent to the controller through the communication interface; and the operation parameter is adjusted by the controller.

[0125] It should be noted that, by means of the online monitoring and real-time correction mechanism, the system operation state deviation can be found in time and corrected by collecting the operation parameters in real time, comparing with the optimized parameters, and adjusting by using the PID algorithm, so that the grid-connected VRFB energy storage system can be ensured to operate stably and efficiently.

[0126] In the present application, the operation parameters of the grid-connected VRFB energy storage system are collected in real time by sensors, and the specific arrangement of the sensors is as follows:

[0127] The sensors are arranged at key positions of the grid-connected VRFB energy storage system.

[0128] Current and voltage sensors are arranged in the VRFB stack to collect voltage and current data during the charging and discharging process of the stack, and the state of charge of the electrolyte is calculated in real time by using the ampere-hour integration method and the open circuit voltage method; it should be noted that the ampere-hour integration method and the open circuit voltage method for calculating the state of charge are conventional settings, and will not be described here;

[0129] Electrolyte temperature sensors are distributed in the stack and the electrolyte storage container to monitor the temperature change of the electrolyte.

[0130] Voltage sensors are installed at the output end of the stack, the input and output ends of the inverter, and the grid connection point to measure the voltage value.

[0131] Current sensors are arranged on the stack and the inverter to monitor the charging and discharging current of the stack and the output current of the inverter.

[0132] The operation parameters collected by the sensors are processed by the data acquisition module to convert the analog signals into digital signals, and are transmitted to the monitoring center by wired or wireless communication.

[0133] It should be noted that the sensor deployment scheme provides a hardware foundation for the stability and economy of the system by accurately covering the key positions and cooperatively monitoring multiple parameters, so that the state of the grid-connected VRFB energy storage system in actual operation is known, controllable and optimizable.

[0134] An apparatus includes a memory and a processor, the memory stores a computer program of the above parameter optimization method, and the processor executes the program to realize the parameter optimization of the grid-connected VRFB energy storage system.

[0135] It should be noted that by converting the parameter optimization method into an executable computer program, the device provides a standardized and intelligent solution for the engineering application of the grid-connected VRFB energy storage system, and promotes the development of energy storage technology in the direction of high reliability and high adaptability.

[0136] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the specification and examples be considered as exemplary only, with the true scope and spirit of the application being indicated by the following claims.

[0137] It is to be understood that the application is not limited to the precise construction here described and illustrated in the accompanying drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application should only be limited by the appended claims.

Claims

1. A parameter optimization method for a network-constructed VRFB energy storage system, characterized in that, Comprising the following steps: S1, constructing an extended state space model: Based on the physical architecture and electrical characteristics of the VRFB energy storage system, an extended state space model containing structural parameters and dynamic operating states is established: Obtain the structural parameters and operating state parameters of the VRFB energy storage system and convert them into a state vector x; wherein the state vector x includes the inverter output angular frequency w, the grid-connected point active power P, the grid-connected point reactive power Q, the inverter output current idq in the dq coordinate system, the filtered output voltage vdq in the dq coordinate system, the voltage loop differential operator x1, x2, the current loop differential operator x3, x4, and the DC voltage Udc; Establishing state equations and output equations ; where A is the state matrix containing the stack dynamics, inverter control logic and grid coupling terms, B is the input matrix, C is the output matrix, D is the feed forward matrix, u is the input vector, y is the output vector; Substitute the state vector x and the input vector u into the state equation and the output equation to obtain the extended state space model; S2, eigenvalue calculation and stability criterion: Substitute the engineering parameters to solve the eigenvalues of the state matrix, and judge whether there are eigenvalues that cause the system to diverge, resonate or lack robustness, as follows: Obtain the engineering parameters; the engineering parameters include VRFB stack physical parameters, inverter control parameters, filter parameters, and thermodynamic parameters; the VRFB stack physical parameters include single stack capacity, internal resistance, electrolyte temperature T, Faraday constant F, battery state of charge SOC, gas constant R, number of battery plates in the stack, number of stack series, number of stack parallel, battery standard electrode potential E; inverter control parameters include virtual moment of inertia G, damping coefficient D, virtual impedance control parameters; filter parameters include inductance value L1 on the inverter side, inductance value L2 on the grid side, filter capacitance value Cf, dq coordinate voltage and current static values Vdq0 output by the filter, grid angular frequency W, and filter capacitance value on the DC side; thermodynamic parameters include specific heat capacity of electrolyte cp and heat dissipation coefficient h; Substitute the engineering parameters into the state matrix A, solve the characteristic equation , and obtain the eigenvalues of the state matrix , where The imaginary part is used to determine the system stability The real part is used to reflect the oscillation characteristics According to the real part and the imaginary part of the eigenvalue, judge whether the system has divergence, resonance or lack of robustness, as follows: If the real part Re(λ) of the eigenvalue is greater than 0, ), the system is unstable and may lead to divergence. Obtain a certain number of continuous eigenvalues, and calculate the real part difference by subtracting the real part of the certain number of eigenvalues from 0; Calculate the standard deviation of all real part differences to obtain the real part trend value; Weight the current real part difference and the real part trend value to obtain the real part evaluation value; if the real part evaluation value is less than the set proximity threshold, it indicates that the real part is close to 0; If there exists an imaginary part Im of the eigenvalues ​​( If the eigenvalues ​​are not equal to 0 and the real part is close to 0, it indicates that the system may resonate; if the distribution of eigenvalues ​​is too scattered or close to the imaginary axis, it indicates that the robustness of the system may be insufficient. S3, multi-objective optimization function construction: Based on the real part, imaginary part and damping ratio of the eigenvalue, an optimization target is constructed, and an intelligent algorithm is used to solve the optimal parameter combination; S4, optimization result verification and correction: Verify the effectiveness of the parameters through multi-working-condition closed-loop simulation, and trigger parameter correction if the indicators are not met.

2. The parameter optimization method of a network-constructed VRFB energy storage system according to claim 1, characterized in that, Based on the real part, imaginary part and damping ratio of the eigenvalue, an optimization target is constructed, and an intelligent algorithm is used to solve the optimal parameter combination, as follows: When the intelligent algorithm uses genetic algorithm for optimization, the specific optimization steps are as follows: S31, set the algorithm parameters of the genetic algorithm, including population size N, iteration number T1, crossover probability pc, and mutation probability pm; S32, encode the parameter combination to be optimized in binary, identify the number of parameters to be optimized as m, and represent each parameter with a k-bit binary number, so the encoding length of each individual is L=m*k bits; S33, randomly generate N individuals to form an initial population; each individual corresponds to a set of parameter combinations, and the specific parameter values are obtained by decoding; S34, for each individual in the population, substitute the corresponding parameter combination into the state matrix, solve the characteristic equation to obtain the eigenvalue, and then calculate the fitness value according to the multi-objective optimization function, the calculation formula is: + + where w1, w2, w3 are weight coefficients, and w1+w2+w3=1, n is the number of eigenvalues of the state matrix, is the real part of the i-th eigenvalue, and the sum of squares of the real parts reflects the stability of the system, is the damping ratio corresponding to the i-th eigenvalue, ; S35, select individuals from the parent generation for crossover operation according to the crossover probability pc; S36, perform mutation operation on the newly generated child individuals according to the mutation probability pm; S37, determine whether the termination condition is met, which includes reaching the number of iterations and the improvement amplitude being less than the improvement threshold; if the termination condition is met, the algorithm stops, and the parameter combination corresponding to the individual with the smallest fitness value in the current population is output as the optimal parameter combination; otherwise, return to step S34 to continue iteration.

3. The parameter optimization method of a network-constructed VRFB energy storage system according to claim 2, characterized in that, Verify the effectiveness of the parameters through multi-condition closed-loop simulation, and trigger parameter correction if the indicators are not met, as follows: Condition setting and simulation execution: Perform multi-condition closed-loop simulation using the optimized parameters, set the condition, including load mutation and power grid failure; Substitute the optimized parameters into the network VRFB energy storage system mathematical model containing the VRFB stack model, inverter model, controller model, and power grid model for closed-loop simulation; Performance index monitoring: During the simulation process, real-time monitoring of various performance indicators of the system is performed, including voltage stability, frequency stability, and power tracking accuracy; for each performance indicator, a corresponding threshold is set, including the DC voltage deviation threshold, the frequency deviation threshold, and the power tracking error threshold; If any indicator is greater than its corresponding threshold, it is determined that the indicator does not meet the requirements: When the voltage stability or frequency stability is greater than its corresponding threshold, increase the real part weight of the eigenvalue in the multi-objective optimization function that is strongly related to system stability, and decrease the crossover probability of the genetic algorithm according to the set rules; If the power tracking accuracy is greater than its corresponding threshold, increase the imaginary part weight of the eigenvalue in the multi-objective optimization function that is related to system dynamic characteristics, and increase the mutation probability according to the set rules; Use the adjusted weight coefficients and intelligent algorithm parameters to re-run the genetic algorithm for optimization and solution, obtain a new parameter combination, and then perform multi-condition closed-loop simulation verification again, repeat the above steps until all performance indicators meet the requirements. In the optimization result verification and correction step, there is also an online monitoring and real-time correction mechanism applied in the monitoring center:

4. The parameter optimization method of a network-constructed VRFB energy storage system according to claim 1, characterized in that, During the actual operation of the system, the running parameters of the network VRFB energy storage system are collected in real time through sensors; After the system is optimized, the optimized running parameters are stored in the database of the monitoring center; Compare the running parameters with the optimized running parameters and calculate the deviation value; ​ The time region within the set running monitoring time length before the current time is recorded as a running monitoring time zone; the deviation value at any collection time within the running monitoring time zone is identified, and the deviation trend value and the deviation average value are obtained by calculating the deviation values within the running monitoring time zone using the standard deviation and average formula; the deviation value, the deviation trend value, and the deviation average value at the current time are weighted to obtain the deviation evaluation value; A corresponding threshold value is set for the deviation evaluation value of each running parameter, and if the deviation evaluation value of the running parameter is greater than the corresponding threshold value, it indicates that the system running state deviates from the optimal state, and the real-time correction mechanism needs to be triggered, and an adjustment instruction is generated; The real-time correction mechanism adopts a PID control algorithm: The running parameter corresponding to the trigger real-time correction mechanism is denoted as u, and the current deviation value The output of the PID algorithm is the parameter adjustment amount The calculation formula of the PID control algorithm is Wherein are the proportional coefficient, integral coefficient and differential coefficient, respectively The adjustment instruction and the parameter adjustment amount are sent to the controller through the communication interface; and the running parameter is adjusted by the controller.

5. The parameter optimization method of a network-constructed VRFB energy storage system according to claim 4, characterized in that, The running parameters of the network-type VRFB energy storage system are collected in real time by sensors, and the specific settings of the sensors are as follows: The sensors are deployed at key positions of the network-type VRFB energy storage system: Current and voltage sensors are arranged in the VRFB stack to collect voltage and current data during the charging and discharging process of the stack, and the state of charge of the electrolyte is calculated in real time by using the ampere-hour integration method and the open circuit voltage method; Electrolyte temperature sensors are distributed in the stack and the electrolyte storage container to monitor the temperature change of the electrolyte; Voltage sensors are installed at the output end of the stack, the input and output ends of the inverter, and the grid connection point to measure the voltage value; Current sensors are arranged on the stack and the inverter to monitor the charging and discharging current of the stack and the output current of the inverter; The running parameters collected by the sensors are processed by the data acquisition module to convert the analog signals into digital signals, and are transmitted to the monitoring center through wired or wireless communication.

6. A parameter optimization device for a network-constructed VRFB energy storage system, characterized by, The device comprises a memory and a processor, the memory stores a computer program of the parameter optimization method according to any one of claims 1-5, and the processor executes the program to realize the parameter optimization of the network-type VRFB energy storage system.

Images