A method for controlling AC and DC faults of photovoltaic energy storage systems based on virtual synchronous machines

Abstract

The present application provides a method for controlling faults of a multi-terminal AC / DC system with photovoltaic storage based on a virtual synchronous machine, which relates to the field of power distribution optimization technology, including: when the total load power meets the critical power threshold, according to the dynamic power distribution ratio, the target charging station is coordinatedly distributed using the distribution network and the photovoltaic system; based on the distribution twin network, distribution faults are predicted according to the status monitoring data to generate fault prediction information; flexible distribution analysis is performed to generate a distribution coordination strategy; when a distribution fault is detected, the target charging station is controlled based on the virtual synchronous machine and the distribution coordination strategy. This application can solve the technical problem of the poor stability and reliability of the vehicle charging service caused by the existing power distribution method of a multi-terminal AC / DC system with photovoltaic storage for automobile charging stations due to the inability to quickly respond to emergency distribution faults and to perform targeted distribution coordination control. The continuity and reliability of the charging service can be guaranteed, and the safety of the system operation can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of power distribution optimization, and in particular to a method for controlling multi-terminal AC / DC faults of photovoltaic storage devices based on a virtual synchronous machine. Background Art

[0002] With the increasing popularity of electric vehicles, the demand for car charging stations is also growing. In order to meet this demand, the photovoltaic storage multi-terminal AC / DC system, as a new type of energy management system, is widely used in car charging stations. This system combines photovoltaic power generation, energy storage system and AC / DC conversion technology, which can effectively utilize renewable energy and provide stable and reliable charging services for electric vehicles. However, the existing photovoltaic storage multi-terminal AC / DC system distribution method has certain limitations when dealing with emergency distribution failures.

[0003] At present, the existing photovoltaic storage multi-terminal AC / DC system distribution method used in automobile charging stations has technical problems such as poor stability and reliability of automobile charging services due to the inability to quickly respond to emergency distribution failures and to perform targeted distribution coordination control. Summary of the invention

[0004] The purpose of this application is to provide a photovoltaic storage multi-terminal AC / DC fault control method based on a virtual synchronous machine to solve the technical problem that the stability and reliability of the vehicle charging service are poor due to the inability to quickly respond to emergency distribution faults and to carry out targeted distribution coordination control in the existing photovoltaic storage multi-terminal AC / DC system distribution method used for vehicle charging stations.

[0005] In view of the above problems, the present application provides a photovoltaic storage multi-terminal AC / DC fault control method based on a virtual synchronous machine, including: executing the distribution operation of the target charging station through the photovoltaic system, and obtaining the total load power in real time, wherein the photovoltaic system has a first distribution priority; fixed-point monitoring of the power generation factor of the photovoltaic system, predicting the power generation power based on the power generation factor, setting a critical power threshold, and when the total load power meets the critical power threshold, according to the dynamic distribution ratio, using the distribution network and the photovoltaic system to coordinate the distribution of power to the target charging station, wherein the distribution network has a second distribution priority; according to predetermined monitoring indicators, the distribution status of the distribution network and the photovoltaic system are respectively monitored, and based on the distribution twin network, distribution fault prediction is performed according to the status monitoring data to generate fault prediction information; based on the fault prediction information, flexible distribution analysis is performed to generate a distribution coordination strategy, wherein the distribution coordination strategy is a distribution plan that includes an energy storage system; when a distribution fault is detected, flexible distribution control is performed on the target charging station based on the virtual synchronous machine and the distribution coordination strategy.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The power generation factor based on the monitoring photovoltaic system is used to predict the power generation, and the critical power threshold is set according to the predicted power generation; then the total load power of the target charging station is monitored in real time. When the total load power exceeds the critical power threshold, the target charging station is coordinated with the distribution network and the photovoltaic system according to the dynamic distribution ratio, wherein the photovoltaic system has the first distribution priority and the distribution network has the second distribution priority; further, in the coordinated distribution process, the distribution network and the photovoltaic system are respectively monitored according to the predetermined monitoring indicators to obtain the status monitoring data; then, based on the distribution twin network, the distribution fault is predicted according to the status monitoring data to generate fault prediction information; then, based on the fault prediction information, the energy storage system is combined to perform flexible distribution analysis and generate a distribution coordination strategy; finally, when a distribution fault is detected, the target charging station is controlled by a flexible distribution based on the virtual synchronous machine and the distribution coordination strategy. Timely response to distribution emergency faults can be achieved, and targeted distribution coordination strategies can be generated for distribution emergency faults, which can improve the accuracy and adaptability of the distribution coordination strategy setting, ensure the continuity and reliability of charging services, and improve the stability and safety of system operation.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0010] Figure 1 This is a flow chart of a method for controlling multi-terminal AC / DC faults of a photovoltaic storage system based on a virtual synchronous machine in the present application;

[0011] Figure 2 This is a flow chart of setting a critical power threshold in a method for controlling a multi-terminal AC / DC fault of an optical storage system based on a virtual synchronous machine in the present application. DETAILED DESCRIPTION

[0012] This application provides a method for controlling faults of photovoltaic and multi-terminal AC and DC systems based on a virtual synchronous machine, which solves the technical problem that the stability and reliability of automobile charging services are poor due to the inability to quickly respond to emergency distribution faults and to carry out targeted distribution coordination control. Timely response to emergency distribution faults can be achieved, and targeted distribution coordination strategies can be generated for emergency distribution faults, which can improve the accuracy and adaptability of the distribution coordination strategy settings, ensure the continuity and reliability of charging services, and improve the stability and safety of system operation.

[0013] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0014] For examples, please see the attached Figure 1 The present application provides a method for controlling a multi-terminal AC / DC fault of an optical storage system based on a virtual synchronous machine, which specifically includes the following steps:

[0015] Step 1: Execute the power distribution operation of the target charging station through the photovoltaic system, and obtain the total load power in real time, wherein the photovoltaic system has the first power distribution priority.

[0016] Specifically, the photovoltaic multi-terminal AC / DC system is a power system that combines photovoltaic power generation, energy storage system and AC / DC conversion technology, which can improve energy utilization efficiency, reduce operating costs and achieve sustainable charging. First, set the first priority of the photovoltaic system in the power distribution process, that is, on the premise of meeting the total power demand of the load, give priority to using the photovoltaic power generation system for power supply, and at the same time adjust the output power of the photovoltaic system according to the total power demand of the load to meet the needs of the charging station. Then, the total load power of the target charging station is monitored in real time, where the target charging station is a car charging station, which includes multiple different types of car charging piles. Through the sensors and monitoring equipment installed on the charging station, the charging power of all charging piles in the charging station is monitored in real time to obtain the total load power. The total load power is the sum of the charging powers of multiple charging piles in the target charging station.

[0017] Step 2: Monitor the power generation factor of the photovoltaic system at a fixed point, predict the power generation based on the power generation factor, set a critical power threshold, and when the total load power meets the critical power threshold, use the distribution network and the photovoltaic system to coordinate power distribution to the target charging station according to the dynamic power distribution ratio, where the distribution network has the second distribution priority.

[0018] Specifically, at a predetermined time node, the key power generation factors of the photovoltaic system, such as light intensity, light angle, temperature, etc., are collected in real time through sensors and monitoring equipment installed on the photovoltaic system to obtain real-time power generation factors; then a power prediction channel is constructed based on the BP neural network, and the real-time power generation factors are input into the power prediction channel to predict the power generation of the photovoltaic system and obtain the predicted power generation power; further, a critical power threshold is set according to the predicted power generation power. This threshold indicates that when the total load power is higher than this value, other power sources need to be called to meet the power demand of the charging station.

[0019] The total load power is judged according to the critical power threshold. When the total load power is greater than the critical power threshold, the dynamic power distribution ratio is obtained. The dynamic power distribution ratio includes the power distribution ratio of the distribution network and the photovoltaic system. It is dynamically set according to the predicted load power, that is, under the preset adjustment node, the dynamic power distribution ratio can be optimized and adjusted according to the actual operation and prediction results to meet the load demand. Then, according to the dynamic power distribution ratio, the distribution network and the photovoltaic system are used to coordinate power distribution to the target charging station, wherein the distribution network has a second power distribution priority, that is, during coordinated power distribution, the photovoltaic system is the first power distribution priority, and renewable energy is used for power supply first. The distribution network is the second power distribution priority, and when photovoltaic power generation is insufficient to meet the power demand of the charging station, it intervenes to supply power. By using the distribution network and the photovoltaic system to coordinate power distribution to the target charging station according to the dynamic power distribution ratio, the electric vehicle charging station can realize the flexible allocation of power demand, improve the energy efficiency and sustainability of the system, and reduce operating costs at the same time, providing users with more economical and reliable charging services.

[0020] Step 3: Monitor the distribution status of the distribution network and photovoltaic system respectively according to the predetermined monitoring indicators, and based on the distribution twin network, predict distribution faults according to the status monitoring data to generate fault prediction information.

[0021] Specifically, the predetermined monitoring indicators are obtained, wherein the predetermined monitoring indicators refer to the distribution operation characteristics with a high degree of correlation with the distribution fault, including distribution monitoring indicators and photovoltaic monitoring indicators, such as distribution monitoring indicators including voltage, current, frequency, etc., and photovoltaic monitoring indicators including light intensity, temperature, component efficiency, etc. Then, based on the predetermined monitoring indicators, the operation data of the distribution system and the photovoltaic system are collected in real time through sensors and monitoring equipment to obtain the status monitoring data. On the other hand, based on the digital twin technology, the distribution network and the photovoltaic system are simulated and modeled respectively. The digital twin technology can simulate and predict the operation status of the system in real time by creating a digital copy of the physical system to generate a distribution twin network; then the status monitoring data is input into the distribution twin network to predict the distribution fault within the preset time window, wherein the preset time window can be set according to the actual situation, for example: preset the distribution fault within the next 20 minutes, and obtain the fault prediction information, wherein the fault prediction information includes the predicted fault type, the predicted fault duration and the predicted fault range, the predicted fault type includes voltage fluctuation, current anomaly, equipment failure, etc., and the predicted fault range includes the number of charging piles affected, the affected area, etc.

[0022] Step 4: Perform flexible power distribution analysis based on the fault prediction information and generate a power distribution coordination strategy, wherein the power distribution coordination strategy is a power distribution plan that includes an energy storage system.

[0023] Specifically, first, the availability and power supply capacity of resources such as photovoltaic power generation systems, energy storage systems, and power grids are evaluated to obtain the distribution resource evaluation results; then, based on the fault prediction information and resource evaluation results, the distribution coordination strategy is optimized to determine the best distribution plan, namely the distribution coordination strategy. The optimal distribution plan includes the distribution type, distribution ratio, energy storage distribution power, and energy storage distribution duration. The distribution types include photovoltaic distribution, power grid distribution, and energy storage distribution; the distribution ratio includes photovoltaic ratio, power grid ratio, and energy storage ratio.

[0024] Step 5: When a power distribution fault is detected, flexible power distribution control is performed on the target charging station based on the virtual synchronous machine and the power distribution coordination strategy.

[0025] Specifically, multiple sensors are used to monitor power distribution faults in real time. First, multiple sensors are deployed at key locations of the charging station to monitor key parameters such as voltage, current, frequency, and temperature in real time to ensure that the sensors can accurately reflect the operating status of the system and provide reliable fault detection data. Then, the sensor data is monitored in real time, and the data is analyzed using advanced algorithms and models to detect potential power distribution faults. When a fault is detected, the system should be able to respond quickly and trigger the fault handling process. When a power distribution fault is detected, flexible power distribution control is performed on the target charging station based on the virtual synchronous machine and the power distribution coordination strategy. First, the output of the virtual synchronous machine is adjusted through the control strategy to simulate the operating characteristics of the synchronous generator and provide support for the grid frequency and voltage. Then, according to the fault type and impact range, the power distribution coordination strategy is executed. When executing the power distribution coordination strategy, the system should be able to flexibly adjust the output power of the power supply and the operating status of the load to adapt to the fault situation. Flexible power distribution control helps to respond to faults quickly, reduce the impact of faults on the operation of charging stations, improve the stability and reliability of the system, and ensure the continuity and reliability of charging services.

[0026] By providing a photovoltaic storage multi-terminal AC / DC fault control method based on a virtual synchronous machine, the technical problem of poor stability and reliability of vehicle charging services caused by the inability to quickly respond to power distribution emergency faults and to carry out targeted power distribution coordination control in the existing photovoltaic storage multi-terminal AC / DC system distribution method used in vehicle charging stations can be solved. First, the power distribution operation of the target charging station is performed through the photovoltaic system, and the total load power is obtained by real-time monitoring, wherein the photovoltaic system has a first distribution priority; then, the power generation factor of the photovoltaic system is monitored at a fixed point, the power generation power is predicted based on the power generation factor, and a critical power threshold is set. When the total load power meets the critical power threshold, the distribution network and the photovoltaic system are used to coordinate power distribution to the target charging station according to the dynamic distribution ratio, wherein the distribution network has a second distribution priority; then, the distribution status of the distribution network and the photovoltaic system are monitored respectively according to predetermined monitoring indicators, and distribution faults are predicted based on the status monitoring data based on the distribution twin network to generate fault prediction information; next, flexible distribution analysis is performed based on the fault prediction information to generate a distribution coordination strategy, wherein the distribution coordination strategy is a distribution plan that includes an energy storage system; finally, when a distribution fault is detected, flexible distribution control is performed on the target charging station based on the virtual synchronous machine and the distribution coordination strategy. It can achieve timely response to power distribution emergency failures and generate targeted power distribution coordination strategies for power distribution emergency failures, which can improve the accuracy and adaptability of power distribution coordination strategy settings, ensure the continuity and reliability of charging services, and improve the stability and safety of system operation.

[0027] Furthermore, based on the power generation factor, the power generation prediction is performed and the critical power threshold is set, as shown in the attached figure. Figure 2As shown, step 2 of this application includes:

[0028] At a predetermined node, power generation factors are collected according to predetermined monitoring indicators to obtain real-time power generation factors, wherein the predetermined monitoring indicators include light intensity, light angle, ambient temperature and ambient humidity; the real-time power generation factors are input into a power prediction channel, predicted power generation is output, and a critical power threshold is set based on the predicted power generation; a method for constructing a power prediction channel is as follows: taking the power generation attribute information of the photovoltaic system as a constraint, conducting homologous information retrieval online to obtain a sample power generation data set, wherein the sample power generation data includes sample light intensity, sample light angle, sample temperature, sample humidity and sample power generation; the sample power generation data set is divided into Q parts, and selected Q times with replacement to generate a first training set, and iteratively selected Q times to obtain Q training sets; taking the sample light intensity, sample light angle, sample temperature and sample humidity as input, taking the sample power generation as supervision, using the Q training sets to supervise the BP neural network respectively, to obtain Q power prediction units that meet the predetermined convergence constraints, and constructing a power prediction channel based on the Q power prediction units, wherein the output of the power prediction channel is the mean of the output results of the Q power prediction units.

[0029] Specifically, predetermined monitoring indicators are obtained, wherein the predetermined monitoring indicators include light intensity, light angle, ambient temperature and ambient humidity; then, corresponding sensors are installed on the photovoltaic system based on the predetermined monitoring indicators, such as light intensity sensors, angle sensors, temperature sensors and humidity sensors, while ensuring that the positions and installation methods of the sensors can accurately reflect the monitored environmental conditions; then, at a predetermined node, data on light intensity, light angle, ambient temperature and ambient humidity are collected in real time through sensors, and the collected data are preprocessed, including calibration, denoising, data conversion, etc., to improve the quality and availability of the data and obtain real-time power generation factors.

[0030] The method for constructing the power prediction channel is as follows: first, the power generation attribute information of the photovoltaic system is used as the retrieval constraint condition, and the same source information is retrieved online, that is, sample power generation data that meets the retrieval condition is searched in multiple data sources, and the data sources include historical data records, online databases, research papers, industry reports, etc., to obtain a sample power generation data set, wherein the sample power generation data includes sample light intensity, sample light angle, sample temperature, sample humidity and sample power generation. Then, the sample power generation data set is divided into Q parts, and Q parts of the data set are selected with replacement Q times to generate the first training set, and the same method is used to iteratively select Q times to obtain Q training sets.

[0031] An initial power prediction unit is constructed based on a BP neural network, wherein the initial power prediction unit is a neural network model that can be iteratively optimized in machine learning, wherein the input data of the initial power prediction unit is light intensity, light angle, temperature and humidity, and the output data is the generated power; then, the sample light intensity, sample light angle, sample temperature and sample humidity are used as input, and the sample generated power is used as supervision, and the BP neural network is supervised and trained using Q training sets, wherein the gradient descent algorithm and the loss function can be used for supervised training, and Q power prediction units that meet the predetermined convergence constraints are obtained. A power prediction channel is further constructed based on the Q power prediction units, wherein the output of the power prediction channel is the mean of the output results of the Q power prediction units. By constructing a power prediction channel based on Q power prediction units, the prediction error of a single prediction unit can be reduced, and the accuracy and reliability of power prediction can be improved.

[0032] Finally, the real-time power generation factor is input into a power prediction channel, the predicted power generation is output, and a critical power threshold is set based on the predicted power generation.

[0033] Further, based on the predicted power generation, a critical power threshold is set, and the present application further includes the following steps:

[0034] Taking a predetermined time threshold as a constraint, the power generation log of the photovoltaic system is retrieved to obtain multiple historical power generation data, wherein the predetermined time threshold is the time interval between a predetermined node and a next adjacent predetermined node; output power volatility analysis is performed based on the multiple historical power generation data, multiple power fluctuation coefficients are determined, and a fluctuation coefficient mean is calculated; deviation calculations are performed on the multiple power fluctuation coefficients and the fluctuation coefficient mean in turn, a power fluctuation coefficient with a fluctuation coefficient deviation less than a predetermined deviation threshold is selected as a frequent fluctuation coefficient, and a mean calculation is performed on multiple frequent fluctuation coefficients to obtain a frequent fluctuation coefficient mean; a power generation reliability coefficient is determined based on the frequent fluctuation coefficient mean analysis, and the predicted power generation power is compensated according to the power generation reliability coefficient to generate a critical power threshold, wherein the frequent fluctuation coefficient is negatively correlated with the power generation reliability coefficient, and the power generation reliability coefficient is greater than 0 and less than 1.

[0035] Specifically, first, a predetermined time threshold is obtained, and the predetermined time threshold is the time interval between a predetermined node and the next adjacent predetermined node; then, with the predetermined time threshold as a constraint, the power generation log of the photovoltaic system is retrieved to obtain multiple historical power generation data, wherein the historical power generation data is the power generation data at the same historical moment as the predetermined time threshold. Then, based on the multiple historical power generation data, output power volatility analysis is performed, that is, the fluctuation of the power generation in each historical power generation data is analyzed. Statistical methods or time series analysis methods can be used to determine the power fluctuation coefficient of each historical power generation data to reflect the degree of fluctuation of the power generation; at the same time, the power fluctuation coefficient of each historical power generation data is calculated. For example, indicators such as standard deviation and coefficient of variation can be used. The power fluctuation coefficient can quantify the degree of fluctuation of the power generation. The greater the degree of fluctuation, the greater the corresponding fluctuation coefficient; multiple power fluctuation coefficients are determined, and the mean of the multiple power fluctuation coefficients is further calculated to obtain the mean of the fluctuation coefficient.

[0036] Then, the deviations of the multiple power fluctuation coefficients and the fluctuation coefficient means are calculated in turn, and the power fluctuation coefficient whose fluctuation coefficient deviation is less than the predetermined deviation threshold is selected as the frequent fluctuation coefficient, wherein the predetermined deviation threshold can be set according to actual conditions, to obtain multiple frequent fluctuation coefficients; further, the mean of the multiple frequent fluctuation coefficients is calculated to obtain the mean of the frequent fluctuation coefficients.

[0037] The power generation reliability coefficient is determined based on the mean value analysis of the frequent fluctuation coefficient, wherein the frequent fluctuation coefficient is negatively correlated with the power generation reliability coefficient, and the power generation reliability coefficient is greater than 0 and less than 1, that is, the larger the frequent fluctuation coefficient, the smaller the corresponding power generation reliability coefficient; finally, the predicted power generation is compensated according to the power generation reliability coefficient, that is, the product of the power generation reliability coefficient and the predicted power generation is set as the critical power threshold to obtain the critical power threshold. By compensating the predicted power generation according to the power generation reliability coefficient, the authenticity, accuracy and reliability of the critical power threshold setting can be improved.

[0038] Further, to set the dynamic power distribution ratio, step 2 of this application includes:

[0039] Based on the historical charging records of the target charging station, the load information within the predetermined time threshold is predicted to obtain the predicted total load power; the deviation of the critical power threshold is calculated according to the predicted total load power to determine the power deviation; the dynamic power distribution ratio is set according to the power deviation and the critical power threshold.

[0040] Specifically, first, call the historical charging records of the target charging station, including key information such as the charging time and charging power of the charging pile; then predict the load information within a predetermined time threshold based on the historical charging records, that is, predict the total load power in the future time period. For example, time series analysis, machine learning algorithms and other methods can be used to predict the load and obtain the predicted total load power. Then, the deviation of the critical power threshold is calculated based on the predicted total load power to determine the power deviation. The power deviation is the difference between the predicted total load power and the critical power threshold; finally, the dynamic power distribution ratio is set according to the power deviation and the critical power threshold, that is, the power distribution ratio of the distribution network and the power distribution ratio of the photovoltaic system are set according to the power deviation and the critical power threshold.

[0041] Further, to obtain the predetermined monitoring indicators, step three of this application includes:

[0042] Obtain a first monitoring indicator space for the distribution network and a second monitoring indicator space for the photovoltaic system, wherein the first monitoring indicator space includes at least voltage, current, frequency, phase and line loss, and the second monitoring indicator space includes at least light intensity, temperature, component efficiency, output power and battery status; respectively perform correlation analysis on multiple first monitoring indicators in the first monitoring indicator space and distribution network faults, and determine that multiple high-frequency first monitoring indicators are set as predetermined first monitoring indicators; respectively perform correlation analysis on multiple second monitoring indicators in the second monitoring indicator space and photovoltaic system faults, and determine that multiple high-frequency second monitoring indicators are set as predetermined second monitoring indicators, and generate the predetermined monitoring indicator according to the predetermined first monitoring indicator and the predetermined second monitoring indicator.

[0043] Specifically, first, obtain the first monitoring indicator space of the distribution network and the second monitoring indicator space of the photovoltaic system, wherein the first monitoring indicator space includes at least voltage, current, frequency, phase and line loss, and the second monitoring indicator space includes at least light intensity, temperature, component efficiency, output power and battery status. Those skilled in the art may also add monitoring indicators according to actual conditions. Then, respectively, analyze the correlation between the multiple first monitoring indicators and the distribution network fault in the first monitoring indicator space, set the first monitoring indicator with a higher degree of correlation as the high-frequency first monitoring indicator, and determine that the multiple high-frequency first monitoring indicators are set as the predetermined first monitoring indicators. On the other hand, respectively analyze the correlation between the multiple second monitoring indicators and the photovoltaic system fault in the second monitoring indicator space, and determine that the multiple high-frequency second monitoring indicators are set as the predetermined second monitoring indicators. Finally, generate the predetermined monitoring indicator according to the predetermined first monitoring indicator and the predetermined second monitoring indicator.

[0044] Further, to determine a plurality of high-frequency first monitoring indicators, the present application further comprises the following steps:

[0045] Retrieve and obtain multiple sample power distribution fault events, perform cluster analysis on the multiple sample power distribution fault events based on the K-means algorithm, and determine multiple sample event sets; eliminate sample event sets whose number of sample events does not meet a predetermined number threshold, and fuse them to obtain a standard sample event set; perform source tracing analysis on the standard sample event set based on an accident tree, and determine multiple associated indicator sets, wherein the associated indicators in each associated indicator set are identified with an indicator order, and each indicator order contains an indicator weight; perform weighted calculation on the same associated indicators in multiple associated indicator sets according to the indicator weight, and obtain an associated indicator sequence by sorting the weighted calculation results from large to small, and extract the first N associated indicators of the associated indicator sequence as high-frequency first monitoring indicators to obtain multiple high-frequency first monitoring indicators.

[0046] Specifically, multiple sample power distribution fault events are retrieved online, that is, data of multiple sample power distribution fault events are collected, including key information such as the time of occurrence of the fault, the type of fault, the duration of the fault, the scope of the fault, etc.; then, cluster analysis is performed on the multiple sample power distribution fault events based on the K-means algorithm, and the sample power distribution fault events are clustered and analyzed using the K-means algorithm to divide the fault events into K different sets, each set representing a fault mode or feature; then, based on the clustering results of the K-means algorithm, multiple sample event sets are determined, each sample event set containing a group of fault events with similar features.

[0047] Further eliminate the sample event set whose number of sample events is less than the predetermined number threshold, and merge the multiple sample event sets remaining after elimination to obtain a standard sample event set; based on the principle of accident tree analysis, perform source tracing analysis on the standard sample event set to determine multiple sets of related indicators. The accident tree is a logical structure diagram used to show the cause-effect relationship between fault events, wherein the related indicators in each related indicator set are marked with an indicator order, and each indicator order contains an indicator weight. The higher the indicator order, the greater the correlation degree, and the corresponding weight is greater. Then, perform weighted calculation on the same related indicators in multiple related indicator sets according to the indicator weight, and sort the same related indicators from large to small according to the weighted calculation results to generate a related indicator sequence; finally, extract the first N related indicators of the related indicator sequence and set them as high-frequency first monitoring indicators to obtain multiple high-frequency first monitoring indicators, where N can be set according to actual needs.

[0048] Further, to generate fault prediction information, step three of this application includes:

[0049] Based on digital twins, in a three-dimensional simulation platform, simulation modeling is performed on the distribution network and the photovoltaic system respectively to obtain a distribution twin model and a photovoltaic twin model; the first state monitoring data is input into the distribution twin model, and a distribution simulation within a preset time window is executed, and first fault prediction information is output, wherein the first state monitoring data is obtained based on monitoring of a predetermined first monitoring indicator; the second state monitoring data is input into the photovoltaic twin model, and a distribution simulation within a preset time window is executed, and second fault prediction information is output, wherein the second state monitoring data is obtained based on monitoring of a predetermined second monitoring indicator; the fault prediction information is constructed according to the first fault prediction information and the second fault prediction information, wherein the fault prediction information includes predicted fault type, predicted fault duration and predicted fault impact range.

[0050] Specifically, digital twin technology is an emerging technology that achieves real-time monitoring, analysis and optimization of physical assets by creating digital copies of physical assets. Based on digital twins, the distribution network and photovoltaic system are simulated and modeled in a three-dimensional simulation platform to obtain a distribution twin model and a photovoltaic twin model. Commonly used three-dimensional simulation platforms include MATLAB Simulink, PowerFactory, etc. Then, the first state monitoring data is input into the distribution twin model, the distribution simulation within the preset time window is performed, and the first fault prediction information is output, wherein the first state monitoring data is obtained based on the predetermined first monitoring index; the second state monitoring data is input into the photovoltaic twin model, the distribution simulation within the preset time window is performed, and the second fault prediction information is output, wherein the second state monitoring data is obtained based on the predetermined second monitoring index; finally, the fault prediction information is constructed according to the first fault prediction information and the second fault prediction information, wherein the fault prediction information includes the predicted fault type, the predicted fault duration and the predicted fault range.

[0051] Further, to generate a power distribution coordination strategy, step 4 of this application includes:

[0052] Obtain distribution quality evaluation indicators, and construct a distribution strategy evaluation function based on the distribution quality evaluation indicators, wherein the distribution quality evaluation indicators include distribution stability, distribution reliability, distribution continuity and distribution economy; based on the energy storage operation parameters of the energy storage system, with the purpose of eliminating the fault prediction information, optimize the distribution plan according to the distribution strategy evaluation function, and output the optimal distribution plan as the distribution coordination strategy, wherein the optimal distribution plan includes distribution type, distribution ratio, energy storage distribution power and energy storage distribution duration.

[0053] Specifically, the power distribution quality evaluation index is obtained, wherein the power distribution quality evaluation index includes power distribution stability, power distribution reliability, power distribution continuity and power distribution economy; then, a power distribution strategy evaluation function is constructed based on the power distribution quality evaluation index, and the power distribution strategy evaluation function is used to comprehensively evaluate the power distribution plan. The energy storage operation parameters of the energy storage system are obtained, including charging and discharging power, energy storage capacity, charging and discharging efficiency, etc.; then, based on the energy storage operation parameters, the power distribution plan is optimized according to the power distribution strategy evaluation function for the purpose of eliminating the fault prediction information. Commonly used optimization algorithms include particle swarm optimization algorithm, genetic algorithm, etc. The appropriate optimization algorithm can be selected according to the actual situation to obtain the optimal power distribution plan and set it as the power distribution coordination strategy, wherein the optimal power distribution plan includes power distribution type, power distribution ratio, energy storage distribution power and energy storage distribution duration.

[0054] Furthermore, the present application also includes the following steps:

[0055] The expression of the distribution strategy evaluation function is:

[0056] P = v1 log 0.5 WR+v2 log 0.5 KR+v3 log 0.5 SR+v4 log 0.5 ZR; where P is the distribution strategy evaluation coefficient, v1 is the distribution stability weight, WR represents the distribution stability, v2 is the distribution reliability weight, KR represents the distribution reliability, v3 is the distribution continuity weight, SR represents the distribution continuity, v4 is the distribution economy weight, and ZR represents the distribution economy.

[0057] Specifically, in the distribution strategy evaluation function, P is the distribution strategy evaluation coefficient, where the distribution strategy evaluation coefficient represents the better the effect of the distribution scheme; v1 is the distribution stability weight, WR represents the distribution stability, v2 is the distribution reliability weight, KR represents the distribution reliability, v3 is the distribution continuity weight, SR represents the distribution continuity, v4 is the distribution economy weight, ZR represents the distribution economy, where the weight is set based on the importance of the indicator to reflect its importance in the distribution strategy evaluation. The higher the importance of the indicator, the greater the corresponding weight.

[0058] In summary, the optical-storage multi-terminal AC / DC fault control method based on a virtual synchronous machine provided by the present application has the following technical effects:

[0059] The power generation factor based on the monitoring photovoltaic system is used to predict the power generation, and the critical power threshold is set according to the predicted power generation; then the total load power of the target charging station is monitored in real time. When the total load power exceeds the critical power threshold, the target charging station is coordinated with the distribution network and the photovoltaic system according to the dynamic distribution ratio, wherein the photovoltaic system has the first distribution priority and the distribution network has the second distribution priority; further, in the coordinated distribution process, the distribution network and the photovoltaic system are respectively monitored according to the predetermined monitoring indicators to obtain the status monitoring data; then, based on the distribution twin network, the distribution fault is predicted according to the status monitoring data to generate fault prediction information; then, based on the fault prediction information, the energy storage system is combined to perform flexible distribution analysis and generate a distribution coordination strategy; finally, when a distribution fault is detected, the target charging station is controlled by a flexible distribution based on the virtual synchronous machine and the distribution coordination strategy. Timely response to distribution emergency faults can be achieved, and targeted distribution coordination strategies can be generated for distribution emergency faults, which can improve the accuracy and adaptability of the distribution coordination strategy setting, ensure the continuity and reliability of charging services, and improve the stability and safety of system operation.

[0060] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0061] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for controlling multi-terminal AC / DC faults of a photovoltaic storage system based on a virtual synchronous machine, characterized in that: include: The photovoltaic system is used to perform power distribution operations at the target charging station, and the total load power is obtained by real-time monitoring, wherein the photovoltaic system has the first power distribution priority; The power generation factor of the photovoltaic system is monitored at a fixed point, the power generation power is predicted based on the power generation factor, and a critical power threshold is set. When the total load power meets the critical power threshold, the distribution network and the photovoltaic system are used to coordinate power distribution to the target charging station according to the dynamic power distribution ratio, wherein the distribution network has the second power distribution priority; The distribution status of the distribution network and photovoltaic system are monitored according to the predetermined monitoring indicators. Based on the distribution twin network, distribution fault prediction is performed according to the status monitoring data to generate fault prediction information; Performing flexible power distribution analysis based on the fault prediction information to generate a power distribution coordination strategy, wherein the power distribution coordination strategy is a power distribution plan that includes an energy storage system; When a power distribution fault is detected, flexible power distribution control is performed on the target charging station based on the virtual synchronous machine and the power distribution coordination strategy; Power generation prediction based on power generation factors and setting critical power thresholds include: At a predetermined node, power generation factors are collected according to predetermined monitoring indicators to obtain real-time power generation factors, wherein the predetermined monitoring indicators include light intensity, light angle, ambient temperature and ambient humidity; Input the real-time power generation factor into a power prediction channel, output a predicted power generation, and set a critical power threshold based on the predicted power generation; The method of constructing a power prediction channel is as follows: Taking the power generation attribute information of the photovoltaic system as a constraint, the same source information is retrieved online to obtain a sample power generation data set, where the sample power generation data includes sample light intensity, sample light angle, sample temperature, sample humidity and sample power generation; Divide the sample power generation data set into Q equal parts, select Q parts with replacement to generate a first training set, and iterate and select Q parts Q times to obtain Q training sets; Taking sample light intensity, sample light angle, sample temperature and sample humidity as input and sample power generation as supervision, supervised training is performed on the BP neural network using Q training sets respectively, and Q power prediction units that meet the predetermined convergence constraints are obtained, and a power prediction channel is constructed based on the Q power prediction units, wherein the output of the power prediction channel is the mean of the output results of the Q power prediction units; Setting a critical power threshold based on the predicted generated power includes: Retrieving the power generation log of the photovoltaic system with a predetermined time threshold as a constraint to obtain a plurality of historical power generation data, wherein the predetermined time threshold is a time interval between a predetermined node and a next adjacent predetermined node; Perform output power fluctuation analysis based on the multiple historical power generation data, determine multiple power fluctuation coefficients, and calculate the average of the fluctuation coefficients; performing deviation calculations on the multiple power fluctuation coefficients and the fluctuation coefficient mean in sequence, selecting a power fluctuation coefficient whose fluctuation coefficient deviation is less than a predetermined deviation threshold as a frequent fluctuation coefficient, and performing mean calculations on the multiple frequent fluctuation coefficients to obtain a frequent fluctuation coefficient mean; Determine the power generation reliability coefficient based on the mean value analysis of the frequent fluctuation coefficient, compensate the predicted power generation power according to the power generation reliability coefficient, and generate a critical power threshold, wherein the frequent fluctuation coefficient and the power generation reliability coefficient are negatively correlated, and the power generation reliability coefficient is greater than 0 and less than 1; Obtain the scheduled monitoring indicators, including: Obtain a first monitoring indicator space of the distribution network and a second monitoring indicator space of the photovoltaic system, wherein the first monitoring indicator space includes at least voltage, current, frequency, phase and line loss, and the second monitoring indicator space includes at least light intensity, temperature, component efficiency, output power and battery status; Respectively performing correlation analysis on a plurality of first monitoring indicators and distribution network faults in the first monitoring indicator space, and determining that a plurality of high-frequency first monitoring indicators are set as predetermined first monitoring indicators; Respectively performing correlation analysis on multiple second monitoring indicators and photovoltaic system faults in the second monitoring indicator space, determining that multiple high-frequency second monitoring indicators are set as predetermined second monitoring indicators, and generating the predetermined monitoring indicator according to the predetermined first monitoring indicator and the predetermined second monitoring indicator; Identify multiple high-frequency first-line monitoring indicators, including: Retrieving and acquiring a plurality of sample power distribution fault events, performing cluster analysis on the plurality of sample power distribution fault events based on a K-means algorithm, and determining a plurality of sample event sets; Eliminate sample event sets whose number of sample events does not meet a predetermined number threshold, and fuse them to obtain a standard sample event set; Based on the accident tree, the standard sample event set is traced and analyzed to determine a plurality of associated indicator sets, wherein the associated indicator in each associated indicator set is identified with an indicator order, and each indicator order includes an indicator weight; Performing weighted calculation on the same related indicators in multiple related indicator sets according to the indicator weights, sorting the weighted calculation results from large to small to obtain a related indicator sequence, and extracting the first N related indicators of the related indicator sequence as high-frequency first monitoring indicators to obtain multiple high-frequency first monitoring indicators; Generates failure prediction information, including: Based on digital twins, the distribution network and photovoltaic system are simulated and modeled in a three-dimensional simulation platform to obtain the distribution twin model and photovoltaic twin model. Inputting first state monitoring data into the power distribution twin model, executing power distribution simulation within a preset time window, and outputting first fault prediction information, wherein the first state monitoring data is obtained based on monitoring of a predetermined first monitoring indicator; Inputting the second state monitoring data into the photovoltaic twin model, executing the power distribution simulation within the preset time window, and outputting the second fault prediction information, wherein the second state monitoring data is obtained based on the predetermined second monitoring indicator; The fault prediction information is constructed according to the first fault prediction information and the second fault prediction information, wherein the fault prediction information includes a predicted fault type, a predicted fault duration, and a predicted fault impact range.

2. According to the method for controlling multi-terminal AC / DC faults of a photovoltaic storage system based on a virtual synchronous machine according to claim 1, it is characterized in that: Set the dynamic power distribution ratio, including: Based on the historical charging records of the target charging station, load information is predicted within a predetermined time threshold to obtain the predicted total load power; Calculating the deviation of the critical power threshold according to the predicted total load power to determine the power deviation; The dynamic power distribution ratio is set according to the power deviation and the critical power threshold.

3. According to the method for controlling multi-terminal AC / DC faults of a photovoltaic storage system based on a virtual synchronous machine according to claim 1, it is characterized in that: Generate distribution coordination strategies, including: Obtaining power distribution quality evaluation indicators, and constructing a power distribution strategy evaluation function based on the power distribution quality evaluation indicators, wherein the power distribution quality evaluation indicators include power distribution stability, power distribution reliability, power distribution continuity and power distribution economy; Based on the energy storage operation parameters of the energy storage system, with the purpose of eliminating the fault prediction information, the distribution plan is optimized according to the distribution strategy evaluation function, and the optimal distribution plan is output as the distribution coordination strategy, wherein the optimal distribution plan includes the distribution type, distribution ratio, energy storage distribution power and energy storage distribution duration.

4. According to the method for controlling multi-terminal AC / DC faults of a photovoltaic storage system based on a virtual synchronous machine according to claim 3, it is characterized in that: The expression of the distribution strategy evaluation function is: P=v1 log 0.5 WR+v2 log 0.5 KR+v3 log 0.5 SR+v4 log 0.5 ZR; Among them, P is the distribution strategy evaluation coefficient, v1 is the distribution stability weight, WR represents the distribution stability, v2 is the distribution reliability weight, KR represents the distribution reliability, v3 is the distribution continuity weight, SR represents the distribution continuity, v4 is the distribution economy weight, and ZR represents the distribution economy.

Images