A method for integrating photovoltaic storage and charging in residential areas

Abstract

The present invention discloses a photovoltaic storage and charging integrated coordination method for Jupei substations, which relates to the field of energy management technology. The photovoltaic storage and charging integrated coordination method for Jupei substations comprises the following steps: step S100, collecting historical data of energy management data, wherein the energy management data comprises energy demand data, photovoltaic power generation data, energy storage equipment data, charging facility data and system operation data; step S200, pre-processing the collected historical data of energy management data to generate energy processing data; step S300, taking the energy processing data as input data, and constructing a machine learning model with power supply abnormality events corresponding to the energy processing data as output; step S400, training the constructed machine learning model, setting an accuracy threshold, and stopping the training when the prediction accuracy of the machine learning model reaches the accuracy threshold; step S500, collecting real-time data of energy management data, and taking the real-time data of energy management data as input.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management, and specifically to an integrated photovoltaic storage and charging coordination method for residential distribution areas. Background Art

[0002] The integrated photovoltaic, energy storage and charging coordination method in the Jupei District is mainly in residential areas, including residential buildings, commercial buildings and other service facilities. It is a coordinated method of energy system built using solar photovoltaic power generation systems, energy storage equipment and electric vehicle charging equipment. Different energy devices can achieve information exchange and coordinated operation, making the energy system more intelligent.

[0003] After searching, the Chinese patent with publication number CN111900727B proposes a PSO-based collaborative optimization scheduling method and device for an integrated photovoltaic, storage and charging station. This technology is based on an integrated photovoltaic, storage and charging station, and performs collaborative optimization scheduling of the output of each module in the charging station according to the collaborative control characteristics of each module during operation, effectively reducing the impact of the integrated charging station on the stability of the power grid system, while also ensuring the safe and stable operation of the integrated charging station itself. In addition, this technology reduces the operating cost of the integrated charging station when providing charging and swapping services to electric vehicle users, while ensuring the effective acceptance of photovoltaic power generation and ensuring full absorption, as well as meeting the normal charging and swapping needs of electric vehicle users, thereby improving its practical value.

[0004] It can be seen from existing technologies that existing intelligent scheduling algorithms have made certain progress in the coordinated scheduling of photovoltaic power generation, energy storage and charging facilities, but there is still room for improvement in accuracy and efficiency, especially considering the diversity and dynamics of energy demand in residential areas. It is necessary to ensure the safety and reliability of the system in the integrated regulation of photovoltaic and energy storage to prevent potential risks and failures. Summary of the invention

[0005] The purpose of the present invention is to provide a method for integrating photovoltaic storage and charging in residential areas to solve the shortcomings of the background technology.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A method for integrating photovoltaic storage and charging in residential areas includes the following steps:

[0008] Step S100, collecting historical data of energy management data, wherein the energy management data includes energy demand data, photovoltaic power generation data, energy storage equipment data, charging facility data and system operation data;

[0009] Step S200, pre-processing the collected historical data of energy management data to generate energy processing data;

[0010] Step S300, using energy processing data as input data and power supply abnormality events corresponding to the energy processing data as output to build a machine learning model;

[0011] Step S400, training the constructed machine learning model, setting an accuracy threshold, and stopping the training when the prediction accuracy of the machine learning model reaches the accuracy threshold;

[0012] Step S500, collect real-time data of energy management data, and use the real-time data of energy management data as input, and output predicted future power supply abnormal events through a machine learning model.

[0013] Furthermore, the energy demand data is real-time monitored by smart meters and energy management systems for electricity consumption data in residential areas; photovoltaic power generation data is real-time collected by installing photovoltaic power generation system monitoring devices to collect light intensity and photovoltaic panel output power; energy storage equipment data is monitored by equipment monitoring devices to monitor the charging and discharging status, charging and discharging efficiency, capacity and power of energy storage equipment; charging facility data is recorded by monitoring equipment installed on the charging facilities to record charging power, charging cycles, charging facility utilization, total available time of charging facilities and average charging power demand; system operation data is used to record the operating status of the system.

[0014] Furthermore, the energy processing data includes statistical data of electricity consumption at different time granularities such as day, week, and month, power generation data information, statistical information on energy storage device status and energy storage conditions, charging facility utilization and charging demand distribution information, and statistical information on system operation status.

[0015] Furthermore, the preprocessing steps of energy management data are:

[0016] According to the collected energy demand data, outlier detection and processing are performed, unreasonable data are eliminated or corrected, time series analysis is performed on the data, periodic and trend changes are identified, and smoothing is performed to eliminate the influence of noise, data interpolation is performed, missing values ​​are filled, and the continuity and integrity of the data are ensured. According to the pre-processed data, statistical data of electricity consumption at different time granularities such as day, week, and month are generated; quality control is performed on the collected photovoltaic power generation data, abnormal data points are detected and eliminated, and data is smoothed using methods such as sliding windows or moving averages to reduce data noise and fluctuations. Data alignment and normalization are performed to ensure that the data has the same time interval and scale, and power generation data information is generated; outlier detection and cleaning are performed on the collected energy storage device data, abnormal data is eliminated, the changing trend of the charge and discharge state is analyzed, and the charge and discharge state conversion point is identified. Perform status marking, calculate the charging and discharging efficiency, capacity, power and other indicators of the energy storage equipment, and perform normalization processing. Based on the preprocessed data, generate statistical information on the status of the energy storage equipment and energy storage conditions; perform anomaly detection and cleaning on the collected charging facility data, eliminate abnormal data, analyze the changing patterns of charging power and charging time periods, identify charging peak and trough periods, count the frequency of use and charging cycles of charging facilities, analyze user charging behavior, and generate charging facility utilization and charging demand distribution information based on the preprocessed data; perform anomaly detection and cleaning on the collected system operation data, eliminate abnormal data, analyze the changing trend of the system operation status, identify system failures and abnormal conditions, count the duration and frequency of the system operation status, evaluate the stability and reliability of the system, and generate statistical information on the system operation status based on the preprocessed data.

[0017] Furthermore, the defined power supply abnormality events include power grid failure, photovoltaic power generation system failure, energy storage equipment failure, charging facility failure, system overload, short circuit failure, voltage abnormality, frequency abnormality and communication interruption. Based on historical data, the corresponding power supply abnormality events in the energy processing data are identified and marked as positive samples.

[0018] Furthermore, the steps of constructing the machine learning model are:

[0019] Step S301: Take the input features of n groups of energy processing data as the input data of the machine learning model, take the corresponding power supply abnormality events as the output labels of the model, extract and transform the features of the energy processing data so that the machine learning model can understand and process them, and use feature selection, dimensionality reduction and other methods to optimize the input features;

[0020] Step S302: Select a machine learning algorithm, including decision tree, random forest, support vector machine, etc., according to the complexity of the problem and the characteristics of the data, use historical data for model training, use energy processing data as input, and use power supply abnormality events corresponding to the energy processing data as output.

[0021] Step S303: Evaluate the performance of the model and perform optimization through cross-validation and other techniques.

[0022] Further, in step S400, the energy processing data is divided into a test set and a training set, the constructed machine learning model is trained, an accuracy threshold is set, and when the prediction accuracy of the machine learning model reaches the accuracy threshold, the training is stopped;

[0023] Step S401: according to the requirements of the problem and the application scenario, a prediction accuracy threshold β of the machine learning model is set to determine whether the performance of the machine learning model meets the requirements;

[0024] Step S402: Process the β feature, including missing value processing, standardization or normalization processing, etc., to ensure that the machine learning model can effectively utilize this feature.

[0025] Step S403: Use the training set to train the model and evaluate the model performance on the validation set. In each round of training, monitor the prediction accuracy of the model until it reaches the set accuracy threshold β. When the prediction accuracy of the model reaches the set threshold β, stop model training. If the prediction accuracy of the machine learning model reaches the set threshold β, continue training until the required accuracy is reached.

[0026] Furthermore, the output steps of the power supply abnormality event are:

[0027] Step S501: Real-time data in the energy management system is collected, including energy demand data, photovoltaic power generation data, energy storage equipment data, charging facility data and system operation data.

[0028] Step S502: The collected real-time data is subjected to the same preprocessing steps as the historical data to ensure the quality and consistency of the input data.

[0029] Step S503: Input the preprocessed real-time data into the trained machine learning model to predict abnormal power supply events. Based on the abnormal power supply events output by the machine learning model as the prediction results, predict the abnormal power supply events that will occur and take corresponding countermeasures in advance.

[0030] Furthermore, a computer server includes a processor and a memory, wherein the memory stores a computer program that can be called by the processor; the processor is used to execute any one of the integrated photovoltaic storage and charging collaborative methods for residential distribution areas by calling the computer program stored in the memory.

[0031] Furthermore, a computer-readable storage medium stores a rewritable computer program thereon, and when the computer program is run on a computer device, the computer device executes any one of the integrated photovoltaic storage and charging coordination methods for residential distribution areas.

[0032] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0033] By collecting real-time data and using machine learning models for prediction, the present invention can accurately predict future power supply abnormal events, help operators take timely countermeasures, and improve the reliability and stability of the power supply system. In the process of building the machine learning model, a large amount of historical data is learned and analyzed, which can extract key information from complex energy management data and provide decision makers with intelligent decision support, thereby improving decision-making efficiency and accuracy, thereby rationally scheduling energy supply and reserves, optimizing energy utilization efficiency, and reducing energy waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0035] Figure 1 The present invention is a flow chart of a method for integrating photovoltaic storage and charging in residential distribution areas. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] See also Figure 1 As shown, this embodiment proposes a photovoltaic storage and charging integrated coordination method in the residential area, including the following steps:

[0038] Step S100, collecting historical data of energy management data, wherein the energy management data includes energy demand data, photovoltaic power generation data, energy storage equipment data, charging facility data and system operation data;

[0039] The energy demand data is real-time monitored by smart meters and energy management systems for electricity consumption data in residential areas; photovoltaic power generation data is real-time collected by installing photovoltaic power generation system monitoring devices to collect light intensity and photovoltaic panel output power; energy storage equipment data is monitored by equipment monitoring devices to monitor the charging and discharging status, charging and discharging efficiency, capacity and power of energy storage equipment; charging facility data is recorded by monitoring equipment installed on the charging facilities to record charging power, charging cycles, charging facility utilization, total available time of charging facilities and average charging power demand; system operation data is used to record the operating status of the system.

[0040] It should be noted that the energy data is obtained through smart meters and existing energy management systems. Energy demand data is crucial for understanding the energy consumption of residential areas. By recording electricity consumption data in different time periods, peak and trough periods can be identified to help formulate strategies for optimizing energy utilization.

[0041] Photovoltaic power generation data is obtained through light sensors, current / voltage sensors and other devices. Photovoltaic power generation data is the key to predicting solar power generation. Real-time data on light intensity and photovoltaic panel output power helps to establish an accurate power generation model and help optimize energy scheduling;

[0042] Energy storage equipment data reflects the operating status and energy storage conditions of the energy storage system. By real-time monitoring of the status of energy storage equipment, energy storage and release can be effectively managed to improve the energy utilization efficiency of the system.

[0043] Charging facility data reflects the charging demand of electric vehicles, helping the dispatching system optimize the use of charging equipment, improve charging efficiency, reduce charging waiting time, and enhance user experience.

[0044] The system operation data records the operating status of the system and is obtained through the system monitoring device.

[0045] Step S200, pre-processing the collected historical data of energy management data to generate energy processing data;

[0046] The energy processing data includes statistical data of electricity consumption at different time granularities such as day, week, and month, power generation data information, statistical information on energy storage device status and energy storage conditions, charging facility utilization and charging demand distribution information, and statistical information on system operation status.

[0047] The preprocessing steps of energy management data are:

[0048] Based on the collected energy demand data, outlier detection and processing are performed to eliminate or correct unreasonable data. Time series analysis is performed on the data to identify periodic and trend changes, and smoothing is performed to eliminate noise effects. Data interpolation is performed to fill in missing values ​​to ensure data continuity and integrity. Based on the pre-processed data, statistical data on electricity consumption at different time granularities such as day, week, and month are generated.

[0049] Perform quality control on the collected photovoltaic power generation data, detect and remove abnormal data points, smooth the data using methods such as sliding windows or moving averages, reduce data noise and fluctuations, perform data alignment and normalization, ensure that the data has the same time interval and scale, and generate power generation data information;

[0050] The formula for obtaining the power generation data information is:

[0051] Pt=i×Pr×η×[1+αt(T-Tr)];

[0052] Among them, Pt is the power generation data information, i is the light intensity, Pr is the output power of the photovoltaic panel, η is the conversion efficiency of the photovoltaic panel, T is the real-time temperature of the photovoltaic panel, the unit is ℃, Tr is the reference temperature of the photovoltaic panel, the setting range is 25-30℃, αt is the temperature coefficient of the photovoltaic panel, which indicates the degree of influence of temperature on the performance of the photovoltaic panel.

[0053] Perform outlier detection and cleaning on the collected energy storage device data, eliminate abnormal data, analyze the changing trend of the charge and discharge status, identify the charge and discharge status conversion points, mark the status, calculate the charge and discharge efficiency, capacity and power of the energy storage device, and perform normalization processing. Based on the pre-processed data, generate statistical information on the energy storage device status and energy storage situation;

[0054] Perform anomaly detection and cleaning on the collected charging facility data, remove abnormal data, analyze the changing patterns of charging power and charging time periods, identify charging peak and trough periods, count the usage frequency and charging cycle of charging facilities, analyze user charging behavior, and generate charging facility utilization and charging demand distribution information based on pre-processed data;

[0055] The generation formula of the charging facility utilization rate and charging demand distribution information is:

[0056]

[0057] Among them, γ is the utilization rate of charging facilities, Pag is the charging demand distribution information, C(t) represents the charging power for charging in time period t, U(t) is the utilization rate of charging facilities for charging in charging period t, A(t) is the total available time of charging facilities, and P(t) is the average charging power demand in time period t.

[0058] Average charging power demand per hour

[0059] Perform anomaly detection and cleaning on the collected system operation data, eliminate abnormal data, analyze the changing trend of the system operation status, identify system failures and abnormal conditions, count the duration and frequency of the system operation status, evaluate the stability and reliability of the system, and generate statistical information on the system operation status based on the preprocessed data.

[0060] Step S300, using energy processing data as input data and power supply abnormality events corresponding to the energy processing data as output to build a machine learning model;

[0061] Abnormal power supply events are defined, including power grid failure, photovoltaic power generation system failure, energy storage equipment failure, charging facility failure, system overload, short circuit failure, voltage abnormality, frequency abnormality and communication interruption. According to historical data, the corresponding abnormal power supply events in the energy processing data are identified and marked as positive samples.

[0062] The steps for building the machine learning model are:

[0063] Step S301: Take the input features of n groups of energy processing data as the input data of the machine learning model, take the corresponding power supply abnormality events as the output labels of the model, extract and transform the features of the energy processing data so that the machine learning model can understand and process them, and use feature selection, dimensionality reduction and other methods to optimize the input features;

[0064] Step S302: Select a machine learning algorithm, including decision tree, random forest, support vector machine, etc., according to the complexity of the problem and the characteristics of the data, use historical data for model training, use energy processing data as input, and use power supply abnormality events corresponding to the energy processing data as output.

[0065] Step S303: Evaluate the performance of the model and perform optimization through cross-validation and other techniques.

[0066] It should be noted that: Grid failure refers to a failure or interruption of the grid, which results in interruption or instability of power supply, such as a power line failure causing power outages in residential areas;

[0067] The photovoltaic power generation system fails, affecting the stability of solar power generation or output power, such as photovoltaic module damage resulting in a decrease in photovoltaic power generation;

[0068] Energy storage equipment failure refers to the failure of energy storage equipment (such as batteries, energy storage systems), which affects the storage and release of energy. For example, battery damage leads to reduced energy storage efficiency.

[0069] Charging facility failure refers to the failure of charging facilities (such as charging piles and charging stations), which affects the charging service of electric vehicles. For example, the charging pile is damaged and the electric vehicle cannot be charged.

[0070] System overload means that the energy management system carries a load that exceeds its design capacity, causing the power supply system to be unstable or fail. For example, peak power consumption exceeds the system design capacity.

[0071] A short circuit fault is a short circuit in the circuit, which causes excessive current or voltage fluctuations, affecting the quality of power supply. For example, a sudden short circuit in the circuit causes a power outage in a local area.

[0072] Voltage abnormality means that the voltage exceeds the normal range, affecting the normal operation of the equipment, such as large voltage fluctuations causing equipment damage;

[0073] Frequency anomaly means that the grid frequency deviates from the normal range, affecting the stability of the power supply system. For example, the grid frequency anomaly causes unstable operation of equipment.

[0074] Communication interruption refers to the interruption of communication between the energy management system and the monitoring equipment, which leads to failure of data collection and control. For example, the communication between the smart meter and the energy management system is interrupted, and the power consumption cannot be monitored in real time.

[0075] Step S400, dividing the energy processing data into a test set and a training set, training the constructed machine learning model, setting an accuracy threshold, and stopping the training when the prediction accuracy of the machine learning model reaches the accuracy threshold;

[0076] Step S401: according to the requirements of the problem and the application scenario, a prediction accuracy threshold β of the machine learning model is set to determine whether the performance of the machine learning model meets the requirements;

[0077] Step S402: Process the β feature, including missing value processing, standardization or normalization processing, etc., to ensure that the machine learning model can effectively utilize this feature.

[0078] Step S403: Use the training set to train the model and evaluate the model performance on the validation set. In each round of training, monitor the prediction accuracy of the model until it reaches the set accuracy threshold β. When the prediction accuracy of the model reaches the set threshold β, stop model training. If the prediction accuracy of the machine learning model reaches the set threshold β, continue training until the required accuracy is reached.

[0079] Step S404: Use the test set to evaluate the trained model, evaluate its generalization ability and true performance, predict the samples in the test set, and calculate the prediction accuracy and other performance indicators of the model.

[0080] It should be noted that we set the accuracy threshold β to 0.85. During the model training process, we can process the β feature, such as standardizing β to ensure that its value range is within the appropriate range. Then, in each round of training, we monitor the prediction accuracy of the model until it reaches the set accuracy threshold β.

[0081] Step S500, collecting real-time data of energy management data, and using the real-time data of energy management data as input, outputting power supply abnormality events through a machine learning model;

[0082] The output steps of power supply abnormality events are as follows:

[0083] Step S501: Real-time data in the energy management system is collected, including energy demand data, photovoltaic power generation data, energy storage equipment data, charging facility data and system operation data.

[0084] Step S502: The collected real-time data is subjected to the same preprocessing steps as the historical data to ensure the quality and consistency of the input data.

[0085] Step S503: Input the preprocessed real-time data into the trained machine learning model to predict abnormal power supply events. Based on the abnormal power supply events output by the machine learning model as the prediction results, predict the abnormal power supply events that will occur and take corresponding countermeasures in advance.

[0086] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0087] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for integrating photovoltaic storage and charging in residential areas, characterized in that: The method comprises the following steps: Step S100, collecting historical data of energy management data, the energy management data includes energy demand data, photovoltaic power generation data, energy storage equipment data, charging facility data and system operation data; photovoltaic power generation data collects light intensity and photovoltaic panel output power in real time by installing a photovoltaic power generation system monitoring device; the energy processing data includes statistical data of electricity consumption at different time granularities such as day, week and month, power generation data information, statistical information of energy storage equipment status and energy storage situation, charging facility utilization rate and charging demand distribution information, and statistical information of system operation status; Step S200, pre-processing the collected historical data of energy management data to generate energy processing data; Step S300, using energy processing data as input data and power supply abnormality events corresponding to the energy processing data as output to build a machine learning model; Step S400, training the constructed machine learning model, setting an accuracy threshold, and stopping the training when the prediction accuracy of the machine learning model reaches the accuracy threshold; Step S500, collecting real-time data of energy management data, and using the real-time data of energy management data as input, and outputting predicted future power supply abnormal events through a machine learning model; The preprocessing steps of energy management data are: Based on the collected energy demand data, outlier detection and processing are performed to eliminate or correct unreasonable data. Time series analysis is performed on the data to identify periodic and trend changes, and smoothing is performed to eliminate noise effects. Data interpolation is performed to fill in missing values ​​to ensure data continuity and integrity. Based on the pre-processed data, statistical data on electricity consumption at different time granularities such as day, week, and month are generated. Perform quality control on the collected photovoltaic power generation data, detect and remove abnormal data points, smooth the data using methods such as sliding windows or moving averages, reduce data noise and fluctuations, perform data alignment and normalization, ensure that the data has the same time interval and scale, and generate power generation data information; The generation formula of the power generation data information is: Pt=i×Pr×η×[1+αt(T-Tr)]; Among them, Pt is the power generation data information, i is the light intensity, Pr is the output power of the photovoltaic panel, η is the conversion efficiency of the photovoltaic panel, T is the real-time temperature of the photovoltaic panel, the unit is ℃, Tr is the reference temperature of the photovoltaic panel, the setting range is 25-30℃, αt is the temperature coefficient of the photovoltaic panel, which indicates the degree of influence of temperature on the performance of the photovoltaic panel; Perform outlier detection and cleaning on the collected energy storage device data, eliminate abnormal data, analyze the changing trend of the charge and discharge status, identify the charge and discharge status conversion points, mark the status, calculate the charge and discharge efficiency, capacity and power of the energy storage device, and perform normalization processing. Based on the pre-processed data, generate statistical information on the energy storage device status and energy storage situation; Perform anomaly detection and cleaning on the collected charging facility data, remove abnormal data, analyze the changing patterns of charging power and charging time periods, identify charging peak and trough periods, count the usage frequency and charging cycle of charging facilities, analyze user charging behavior, and generate charging facility utilization and charging demand distribution information based on pre-processed data; The generation formula of the charging facility utilization rate and charging demand distribution information is: Wherein, γ is the utilization rate of charging facilities, Pag is the distribution information of charging demand, C(t) represents the charging power for charging in time period t, U(t) is the utilization rate of charging facilities for charging in charging period t, A(t) is the total available time of charging facilities, and P(t) is the average charging power demand in time period t; Perform anomaly detection and cleaning on the collected system operation data, eliminate abnormal data, analyze the changing trend of system operation status, identify system failures and abnormal conditions, count the duration and frequency of system operation status, evaluate the stability and reliability of the system, and generate statistical information on system operation status based on pre-processed data; The steps for building the machine learning model are: Step S301: Take the input features of n groups of energy processing data as the input data of the machine learning model, take the corresponding power supply abnormality events as the output labels of the model, extract and transform the features of the energy processing data so that the machine learning model can understand and process them, and use feature selection, dimensionality reduction and other methods to optimize the input features; Step S302: Select a machine learning algorithm, including decision tree, random forest, support vector machine, etc., according to the complexity of the problem and the characteristics of the data, use historical data for model training, use energy processing data as input, and use power supply abnormality events corresponding to the energy processing data as output; Step S303: Evaluate the performance of the model and perform optimization through cross-validation and other techniques.

2. The method for integrating photovoltaic storage and charging in Jupei area according to claim 1 is characterized in that: The energy demand data is real-time monitored by smart meters and energy management systems for electricity consumption in residential areas; the energy storage equipment data is monitored by equipment monitoring devices for the charging and discharging status, charging and discharging efficiency, capacity and power of the energy storage equipment; the charging facility data is recorded by monitoring equipment installed on the charging facilities for charging power, charging cycles, charging facility utilization, total available time of the charging facilities and average charging power demand; the system operation data is used to record the operating status of the system.

3. The integrated photovoltaic storage and charging coordination method for residential areas according to claim 2 is characterized in that: The defined power supply abnormal events include grid failure, photovoltaic power generation system failure, energy storage equipment failure, charging facility failure, system overload, short circuit failure, voltage abnormality, frequency abnormality and communication interruption. Based on historical data, the corresponding power supply abnormal events in the energy processing data are identified and marked as positive samples.

4. The integrated photovoltaic storage and charging coordination method for Jupei area according to claim 1 is characterized in that: Step S400, dividing the energy processing data into a test set and a training set, training the constructed machine learning model, setting an accuracy threshold, and stopping the training when the prediction accuracy of the machine learning model reaches the accuracy threshold; Step S401: according to the requirements of the problem and the application scenario, a prediction accuracy threshold β of the machine learning model is set to determine whether the performance of the machine learning model meets the requirements; Step S402: Processing the β feature, including missing value processing, standardization or normalization processing, etc., to ensure that the machine learning model can effectively utilize this feature; Step S403: Use the training set to train the model and evaluate the model performance on the validation set. In each round of training, monitor the prediction accuracy of the model until it reaches the set accuracy threshold β. When the prediction accuracy of the model reaches the set threshold β, stop model training. If the prediction accuracy of the machine learning model reaches the set threshold β, continue training until the required accuracy is reached. Step S404: Use the test set to evaluate the trained model, evaluate its generalization ability and actual performance, predict the samples in the test set, and calculate the prediction accuracy and other performance indicators of the model.

5. The integrated photovoltaic storage and charging coordination method for Jupei area according to claim 1 is characterized in that: The output steps of power supply abnormality events are as follows: Step S501: real-time data in the energy management system is collected, including energy demand data, photovoltaic power generation data, energy storage equipment data, charging facility data and system operation data; Step S502: Perform the same preprocessing steps on the collected real-time data as on the historical data to ensure the quality and consistency of the input data; Step S503: Input the preprocessed real-time data into the trained machine learning model to predict power supply abnormality events. Based on the power supply abnormality events output by the machine learning model as prediction results, predict the power supply abnormality events that will occur and take corresponding countermeasures in advance.

6. A computer server, comprising a processor and a memory, characterized in that: The memory stores a computer program that can be called by the processor; the processor calls the computer program stored in the memory to execute any one of the integrated photovoltaic, storage and charging coordination methods in residential areas described in claims 1-5.

7. A computer-readable storage medium having a rewritable computer program stored thereon, characterized in that: When the computer program runs on a computer device, the computer device executes any one of claims 1-5 of the integrated photovoltaic storage and charging collaborative method for residential distribution areas.

Images