Dynamic evaluation and optimization method for power grid side energy storage system site selection

Abstract

The invention relates to the technical field of power grid side energy storage, in particular to a dynamic evaluation and optimization method for power grid side energy storage system site selection, which comprises the following steps of: based on a power grid side demand, executing power grid load data analysis, calculating a difference value between a peak load and a minimum load, comparing with historical load data, identifying a fluctuation mode of the power grid demand, and calculating the power grid side energy storage system site selection. And obtaining a load fluctuation mode pattern. According to the invention, by monitoring and analyzing the change of the power grid load in real time, accurately adjusting the deployment of the energy storage equipment and using the long-short-term memory network to optimize the prediction accuracy, the decision error is significantly reduced, the selection of the position and scale of the energy storage system is more scientific, and by combining the cost, the reliability and the environmental influence, the prediction accuracy is improved. The multi-objective optimization enhances the comprehensiveness of the decision, the integration of the geographic information system optimizes the processing of the spatial data, the adaptability and implementation efficiency of the site selection strategy are improved, and the decision is ensured to be implemented under the definite constraint condition, so that the higher operation efficiency and utilization rate of the energy storage system are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of grid - side energy storage, and particularly to a dynamic evaluation and optimization method for the location selection of grid - side energy storage systems. Background Art

[0002] Grid - side energy storage technology involves installing energy storage devices at key positions in the power system to optimize the operation and management of the power grid. The purpose is to store electrical energy and release it during peak demand periods, thereby balancing the grid load, improving energy utilization efficiency, and enhancing the stability of the power grid. Grid - side energy storage systems can take various forms, such as battery energy storage, pumped - storage hydropower, compressed - air energy storage, etc., which can counteract grid load fluctuations, support the wider application of renewable energy, and provide necessary support for aging power infrastructure.

[0003] Among them, the dynamic evaluation and optimization method for the location selection of grid - side energy storage systems focuses on the development and application of strategies to dynamically evaluate and optimize the location of energy storage devices in the power grid. The purpose is to determine the best location for energy storage device location selection through scientific data analysis and optimization algorithms to maximize cost - effectiveness and improve energy efficiency, considering various factors such as the existing layout of the power grid, energy supply - demand dynamics, cost constraints, and technical feasibility.

[0004] Existing technologies often show insufficient response to real - time grid data changes in managing grid - side energy storage systems. Relying on static analysis, they result in adjustment delays in the face of unforeseen demand changes. For example, when the output of renewable energy changes rapidly, such processing limits the flexibility of energy storage systems, causes inefficient use of resources, and increases the grid load. Traditional technologies ignore the immediate interaction between geographical and grid data when conducting location selection and sizing, and cannot adapt to the dynamic changes in geographical and grid environments, thus increasing the complexity and cost of location selection implementation, and affecting the overall performance and adaptability of grid energy storage systems. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose a dynamic evaluation and optimization method for the location selection of grid - side energy storage systems.

[0006] To achieve the above - mentioned purpose, the present invention adopts the following technical solution: A dynamic evaluation and optimization method for the location selection of grid - side energy storage systems, including the following steps: S1: Based on grid - side demand, perform power grid load data analysis, calculate the difference value between the peak load and the lowest load, compare it with historical load data, identify the fluctuation pattern of grid demand, and obtain a load fluctuation pattern diagram; S2: Based on the load fluctuation pattern diagram, utilize the long short-term memory network, according to the real-time input power grid data, perform real-time simulation tests, and by comparing the simulation results with the current power grid data, optimize the prediction error of the energy storage system location selection, and obtain a real-time simulation calibration model; S3: Based on the real-time simulation calibration model, evaluate the location selection and scale of the power grid side energy storage system, combine cost, reliability and environmental impact factors, perform multi-objective optimization, analyze the energy storage requirements of candidate locations, and determine the optimal location selection and scale, and obtain an optimized location selection decision configuration; S4: Based on the optimized location selection decision configuration, combine with the geographic information system, analyze the mutual influence and restriction conditions between geography and power grid data, and utilize the linear programming algorithm to continuously optimize the location selection decision, and obtain the optimal location selection overview result.

[0007] The improvement of the present invention is that the calculation steps of the difference value are specifically as follows: S111: Based on the power grid side demand, collect the power grid load data within the target time period, screen the highest load and the lowest load of each day, and obtain the daily load extreme value data set; S112: Based on the daily load extreme value data set, adopt the formula:

[0008] Calculate the difference value between the peak load and the lowest load , and obtain the difference value data set, where represents the highest load value monitored within the target time period, represents the lowest load value monitored within the same time period.

[0009] The improvement of the present invention is that the steps for obtaining the load fluctuation pattern diagram are specifically as follows: S121: Collect the current and historical power grid load data, adjust it to the time series format, and obtain the integrated load data; S122: Based on the integrated load data, calculate the fluctuation index with the historical load data, and adopt the formula:

[0010] to obtain the load fluctuation score at time , where represents the current load value at time , represents the average value of the historical load data, represents the standard deviation of the historical load data; S123: Based on the load fluctuation score, identify the abnormal time points beyond the normal fluctuation range, and obtain the load fluctuation pattern diagram.

[0011] The improvements of the present invention are as follows. The execution steps of the real-time simulation test are specifically as follows: S211: Based on the load fluctuation pattern diagram, collect and organize the real-time data of the power grid, and verify whether the data format is consistent with the input requirements of the long short-term memory network to obtain the processed data result; S212: Based on the processed data result, input it into the long short-term memory network for real-time simulation test, simulate the power grid behavior, and obtain the simulation test output result.

[0012] The improvements of the present invention are as follows. The acquisition steps of the real-time simulation calibration model are specifically as follows: S221: Compare the simulation result with the current power grid data, and use the formula:

[0013] Calculate the mean absolute error between the predicted value and the current value , and obtain the prediction error evaluation result, where represents the th predicted data point, represents the th current data point, represents the total number of data points; S222: Based on the prediction error evaluation result, analyze the error source, adjust the network structure and learning parameters of the LSTM, re-perform the simulation test on the power grid data, optimize the prediction error of the energy storage system location selection, and obtain the real-time simulation calibration model.

[0014] The improvements of the present invention are as follows. The evaluation steps of the location selection and scale are specifically as follows: S311: Based on the real-time simulation calibration model, perform real-time simulation on potential energy storage sites, evaluate the geographical location, power grid connection points and infrastructure support degree of each site, and obtain the preliminary site performance evaluation data; S312: Based on the preliminary site performance evaluation data, use the formula:

[0015] Calculate the optimization index of the th site , and obtain the evaluation information of multi-site location selection, where represents the power capacity of the th site, represents the reliability index of the th site, represents the construction and operation and maintenance cost of the th site, is the weight coefficient of is the weight coefficient of is the non-linear adjustment coefficient of S313: Analyze the scale and configuration of the corresponding site based on the evaluation information of the multi-site location selection, and obtain the location selection and scale information.

[0016] The improvement of the present invention is that the specific steps for obtaining the optimized location selection decision configuration are as follows: S321: Combine the cost, reliability, and environmental impact factors, perform multi-objective optimization processing, score each candidate location, and judge the energy storage requirements and potential advantages of each location to obtain the location selection scoring result; S322: Based on the location selection scoring result, determine the location with a high score as the most preferred location selection, and determine the scale of the location selection according to the energy storage requirements to obtain the optimized location selection decision configuration.

[0017] The improvement of the present invention is that the specific steps for obtaining the overview result of the most preferred location selection are as follows: S411: Based on the optimized location selection decision configuration, collect geographic information system and power grid data, including terrain, climate, and power grid capacity factors, to obtain the geographic and power grid data set; S412: Based on the geographic and power grid data set, determine the feasibility of the location selection site and the potential difficulties in power grid access to obtain the feasibility analysis result; S413: Conduct a benefit analysis on the feasibility analysis result, using the formula:

[0018] Calculate the benefit index of the th location selection site , combine the geographic and power grid constraints, continuously optimize the location selection decision, and obtain the overview result of the most preferred location selection; where represents the value of the th evaluation parameter, is the weight of the th evaluation parameter, represents the direct cost of the th location, represents the potential risk adjustment value of the th location, is the total number of parameters.

[0019] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In the present invention, by monitoring and analyzing the changes in the power grid load in real time, the deployment of energy storage devices is accurately adjusted. The long short-term memory network is used to optimize the prediction accuracy, significantly reducing the decision-making error, ensuring that the selection of the location and scale of the energy storage system is more scientific. By combining cost, reliability, and environmental impact, multi-objective optimization enhances the comprehensiveness of decision-making. The integration of geographic information systems optimizes the processing of spatial data, improving the adaptability and implementation efficiency of the site selection strategy, and ensuring the implementation of decisions under clear constraints, thereby achieving higher operating efficiency and resource utilization rate of the energy storage system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is the main step flow chart in the present invention; Figure 2 is the calculation flow chart of the difference value in the present invention; Figure 3 is the acquisition flow chart of the load fluctuation pattern diagram in the present invention; Figure 4 is the execution flow chart of the real-time simulation test in the present invention; Figure 5 is the acquisition flow chart of the real-time simulation calibration model in the present invention; Figure 6 is the evaluation flow chart of the location and scale in the present invention; Figure 7 is the acquisition flow chart of the optimized location decision configuration in the present invention; Figure 8 is the acquisition flow chart of the most preferred location overview result in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0022] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. In addition, in the description of the present invention, "a plurality of" means two or more, unless otherwise specifically defined. Embodiment

[0023] Please refer to Figure 1, the present invention provides a technical solution: a dynamic evaluation and optimization method for the location selection of a grid-side energy storage system, including the following steps: S1: Based on the grid-side demand, perform power grid load data analysis, calculate the difference value between the peak load and the minimum load, compare it with the historical load data, identify the fluctuation pattern of the grid demand, and obtain the load fluctuation pattern diagram; the historical load data refers to the actual measurement values of the grid load over a past period of time, including the power consumption and load demand of each region of the power grid. The data covers the grid load conditions in the past few months or years, and records the power consumption in different time periods (such as day, night, seasonal changes, etc.). For example, the power demand in a certain area is higher in the early morning and evening in winter than during the day, or the load changes during the peak period of air conditioner use in summer. The data will help analyze the load fluctuation law of the power grid; S2: Based on the load fluctuation pattern diagram, use a long short-term memory network, perform real-time simulation tests according to the real-time input grid data, optimize the prediction error of the energy storage system location selection by comparing the simulation results with the current grid data, and obtain a real-time simulation calibration model; comparing the simulation results refers to comparing and analyzing the grid load prediction data generated by the simulation with the real-time obtained actual grid load data to evaluate the accuracy of the prediction. During the simulation process, the predicted load demand in a certain area in the next few hours is 500 MW, but the real-time data of the actual grid load shows 450 MW. The difference between the simulation result and the actual data is 50 MW. This comparison helps optimize the location selection and scale prediction of the energy storage system; the prediction error of the location selection specifically refers to that if the prediction shows that a certain area is the best location selection, but during the actual operation process, the power load fluctuation in this area is large or the grid connection conditions are not ideal, then this deviation is the prediction error; S3: Based on the real-time simulation calibration model, evaluate the location selection and scale of the grid-side energy storage system, combine cost, reliability, and environmental impact factors, perform multi-objective optimization, analyze the energy storage demand of the candidate locations, and determine the optimal location selection and scale to obtain an optimized location selection decision configuration; environmental impact factors refer to the external natural environment or social environment conditions that affect the location selection of the grid-side energy storage system. For example, certain areas are not suitable for energy storage systems due to weather conditions (such as wind power, solar radiation, etc.), or certain areas are restricted by environmental protection areas and energy storage facilities cannot be built in the area. The infrastructure construction in the area (such as the accessibility of power lines, transportation conditions) is also an environmental factor affecting the location selection of the energy storage system; S4: Based on the optimized location selection decision configuration, combine the geographic information system, analyze the mutual influence and restriction conditions between geography and grid data, and use the linear programming algorithm to continuously optimize the location selection decision to obtain the optimal location selection overview result.

[0024] The load fluctuation pattern diagram specifically includes peak identification points, trough identification points, and fluctuation periods. The optimized site selection decision configuration is specifically the economic score, the grid-side energy storage stability rating, and the ecological impact analysis results. The optimal site selection overview results include site coordinates, data integration degree, and planning adaptability.

[0025] Please refer to Figure 2 , and the calculation steps of the difference value are specifically as follows: S111: Based on the grid-side demand, collect the grid load data within the target time period, screen the highest load and the lowest load of each day to obtain the daily load extreme value data set; Determine the time range of the target time period, collect the original data including voltage, current, power, etc., upload it to the grid data management platform through the data transmission channel, and the platform formats and stores the received original data in the database. In the data cleaning link, eliminate the outliers and error data in the data. The outlier detection adopts the threshold method and the time series consistency analysis method, divide the data into daily units and classify and organize them. Then, perform the extraction and analysis of the maximum and minimum values on the organized data. The highest load is obtained by comparing the peaks in the daily data, and the lowest load is determined by analyzing the minimum value in the daily load. Finally, organize and generate the daily load extreme value data set.

[0026] S112: Based on the daily load extreme value data set, use the formula:

[0027] Calculate the difference value between the peak load and the lowest load , to obtain the difference value data set, where represents the highest load value monitored within the target time period, represents the lowest load value monitored within the same time period; The following data is collected, is the highest grid load value monitored on a certain day, and 1000MW is monitored, is the lowest load value monitored on the same day, and 600MW is monitored. According to the formula:

[0028] Calculate the difference value between the peak load and the lowest load is 400MW. This result indicates that the load change range of the grid on this day is 400MW, which is crucial for analyzing the load volatility of the grid on this day and can be used for further load management and prediction analysis.

[0029] Please refer to Figure 3 , and the acquisition steps of the load fluctuation pattern diagram are specifically as follows: S121: Collect the current and historical power grid load data, adjust it into a time series format to obtain the integrated load data; The data collected needs to be recorded through smart meters. The current load data is obtained by real-time monitoring of the load values at each power grid node. The data records are marked with timestamps to ensure the correlation and matching of data at different time points. The historical load data is obtained from the power grid operation records. When organizing the data, duplicate removal and outlier detection are required. Incomplete data and significantly deviated outliers are removed through processing. The organized data is reconstructed into a sequence data format based on time. The time series data needs to include time points, load values, and data source identifiers to ensure subsequent analysis and comparison, and the integrated load data is obtained.

[0030] S122: Based on the integrated load data, calculate the fluctuation index with the historical load data, using the formula:

[0031] Obtain the load fluctuation score at time which reflects the standardized deviation of the current load value relative to the historical load data and is used to identify abnormal fluctuations. Among them, represents the current load value at time which is the power grid load data collected in real-time or recently, represents the average value of the historical load data, and the value is calculated from the long-term historical data and is used to represent the average load level of the power grid over a relatively long period of time, represents the standard deviation of the historical load data, which measures the degree of fluctuation of the historical load data and is used to standardize the fluctuation score of the current load value; At a specific time , the current load is 1200MW, the historical average load is 1000MW, and the standard deviation of the historical load is 200MW. Substitute into the formula for calculation:

[0032] A score of 1 indicates that the current load is above one standard deviation, indicating a relatively high load fluctuation and requiring attention to adjustment or management. This result shows that the power grid has a relatively high load pressure at this time point and specific measures need to be taken to handle the overload situation.

[0033]

[0033] S123: Based on the load fluctuation score, identify the abnormal time points that exceed the normal fluctuation range to obtain the load fluctuation pattern diagram; By analyzing the fluctuation scores at each time point, a threshold for anomaly recognition is set. This threshold is usually calculated based on historical data statistical parameters, such as being set within the range of the mean plus twice the standard deviation. All time points beyond this range are marked as anomaly points. In the specific recognition process, it is necessary to compare the load fluctuation scores one by one, screen out all records higher than the threshold, and after marking the time points, map the time points and the corresponding load values onto the fluctuation pattern graph, display the normal fluctuations as a curve, and mark the anomaly points with a specific color. The finally generated chart serves as the load fluctuation pattern graph.

[0034] Please refer to Figure 4 , the specific execution steps of the real-time simulation test are as follows: S211: Based on the load fluctuation pattern graph, collect and organize the real-time data of the power grid, and verify whether the data format is consistent with the input requirements of the long short-term memory network to obtain the processed data result; First, parse the key node data in the fluctuation pattern, including the load upper limit, lower limit, and fluctuation range, to determine the core characteristics of the real-time data of the power grid. By obtaining the real-time load data, verify whether the data contains the key characteristics defined in the load fluctuation pattern graph, remove the outliers in the data and refill the missing values, use the window smoothing method to denoise the data, and correct the fluctuation deviation of each time node by calculating the fluctuation amplitude. Match the processed real-time power grid data with the input format specifications of the long short-term memory network, including dimension adjustment, data normalization, and time series segmentation, and finally generate the processed data result.

[0035] S212: Based on the processed data result, input it into the long short-term memory network for real-time simulation testing, simulate the power grid behavior, and obtain the simulation test output result; Input the processed data result into the long short-term memory network, set the initial state of the network, including the number of network layers, the number of hidden layer nodes, and the time step. Gradually input the time series data into the network, capture the predicted values output by the network in real time, correct the difference between each step of prediction and the input through the internal loop mechanism of the network, and at the same time record the change trend and fluctuation amplitude of each node in the output sequence. Take the predicted sequence generated by the network as the test result of simulating the power grid behavior.

[0036] Please refer to Figure 5 , the specific steps for obtaining the real-time simulation calibration model are as follows: S221: Compare the simulation result with the current power grid data, and use the formula:

[0037] Calculate the mean absolute error between the predicted value and the current value , and obtain the prediction error evaluation result, where represents the a predicted data point, indicating the th current data point, indicating the total number of data points; In a certain power grid data, the five data points for predicting power load are:

[0038] The actual power load values are:

[0039] Then calculate the absolute error of each data point:

[0040] Next, find the average value of the errors:

[0041] This result shows that the average error of the model prediction is 1.6, which provides a benchmark for subsequent optimization. This result shows that the mean absolute error is 1.6, indicating that the prediction model has a small error on these five data points. However, according to the actual requirements, this result needs to further optimize and adjust the model to improve the accuracy of power load prediction.

[0042] S222: Based on the prediction error evaluation results, analyze the error sources, adjust the network structure and learning parameters of the LSTM, re - simulate and test the power grid data, optimize the prediction error of the energy storage system location selection, and obtain a real - time simulation calibration model; Extract error data from the power grid data simulation, including the difference between the actual output and the predicted output. Analyze the error sources by comparing the error value distribution point by point, and then classify and process the errors, which are mainly classified into three types: data anomaly, model under - fitting, and improper network parameter settings. For data anomalies, correct the data set by removing noise points and re - interpolating and filling. For model under - fitting, adjust by increasing the number of hidden layers and improving the network's non - linear expression ability. At the same time, adjust the learning rate and training batches. Specifically, adjust the learning rate from the original value to a dynamically changing method to improve the convergence speed. Finally, re - load the power grid simulation data on the adjusted network and conduct multiple rounds of tests to detect the model accuracy and error distribution. Finally, optimize the prediction error of the energy storage system location selection and obtain a real - time simulation calibration model.

[0043] Please refer to Figure 6 , the evaluation steps for location selection and scale are specifically as follows: S311: Based on the real - time simulation calibration model, conduct real - time simulation on potential energy storage sites, evaluate the geographical location, grid connection points, and infrastructure support of each site, and obtain preliminary site performance evaluation data; By using Geographic Information System (GIS) to precisely locate the geographical position of the site, and at the same time retrieving the basic data of the grid connection point to read and analyze its voltage level and maximum transmission capacity, the infrastructure support degree is calculated by combining on-site investigation and engineering evaluation reports. The evaluation content includes the access capacity of existing grid facilities, the possibility of upgrading and transformation, and the initial investment estimation. The real-time calibration model parameterizes the above data, standardizes each data item according to the preset weights, and forms a comprehensive performance index set including the basic characteristics of the site and the grid access characteristics. The index generates the preliminary performance evaluation data of the site through comprehensive processing.

[0044] S312: Based on the preliminary performance evaluation data of the site, use the formula:

[0045] Calculate the optimization index of the th site to obtain the evaluation information for multi-site location selection, which is used to measure the optimization degree of the site under given conditions. Among them, represents the power capacity of the th site, reflecting the maximum power output that the site can support, represents the reliability index of the th site, which is used to measure the stability and failure rate of the site during operation, represents the construction and operation and maintenance cost of the th site, including the costs of initial construction and long-term operation, is the weight coefficient, which is used to adjust the influence of power capacity in the calculation of the optimization index, is the weight coefficient, which is used to adjust the influence of reliability in the calculation of the optimization index, is the nonlinear adjustment coefficient, which can change the way the cost factor acts in the overall evaluation; The parameter values of a specific site are: power capacity MW, reliability , cost million yuan, weight coefficient , , nonlinear adjustment coefficient . Substitute into the formula to calculate:

[0046] This result shows that based on the given parameters and weights, the optimization index of this site is 0.003, which reflects the optimization degree of the site after comprehensively considering power capacity, reliability and cost. The higher the value, the better the site. This result will be used to compare different sites and determine the optimal site.

[0047] S313: Analyze the scale and configuration of the corresponding site based on the evaluation information for multi-site location selection, and obtain the location selection and scale information; Specifically analyze the scale and configuration of each site, combine the available resources around the site, grid demand, and historical operation data, evaluate the optimal energy storage scale of each site, call the meteorological and load forecasting models to obtain the power demand fluctuation range of the site, and determine the energy storage capacity configuration required for the site in combination with the energy utilization rate and energy storage technology parameters. The configuration plan is further incorporated into the economic evaluation, and the evaluation content includes unit cost, improvement of power quality, and stability of the regional power grid. After completing the above data integration and comparison, the location selection and scale information is obtained.

[0048] Please refer to Figure 7 , and the steps for obtaining the optimized location selection decision configuration are specifically as follows: S321: Perform multi-objective optimization processing in combination with cost, reliability, and environmental impact factors, score each candidate location, judge the energy storage demand and potential advantages of each location, and obtain the location selection scoring result; Sort out the data of the candidate locations one by one. The specific content includes calling the data table about cost in the geographic information system, obtaining the cost parameter value by calculating the cumulative cost of construction and operation, calculating the average failure rate of the candidate location and taking its reciprocal value to represent the reliability parameter. For environmental impact, refer to the local environmental impact assessment report, extract the pollutant emission and ecological impact scores and perform normalization processing to form the environmental impact index. Perform multi-objective optimization calculation on the sorted data above, score each candidate location item by item, compare the optimization indices, and generate a scoring result including the score and energy storage demand of each location to obtain the location selection scoring result.

[0049] S322: Based on the location selection scoring result, determine the location with a high score as the optimal location selection, and determine the scale of the location selection according to the energy storage demand to obtain the optimized location selection decision configuration; Select the location with a higher score as the priority analysis object. For the location with a higher score, by calling the optimization index, environmental impact score, and energy storage demand data in the location selection scoring item by item, compare the relative advantages of the locations in terms of energy storage capacity, construction cost, and operation benefit, analyze the scale setting range in combination with the actual grid demand, and adjust the scale value according to the existing construction standards to complete the scale setting of the optimal location selection, and generate the optimized location selection decision configuration in combination with the energy storage scale.

[0050] Please refer to Figure 8 , and the steps for obtaining the overview result of the optimal location selection are specifically as follows: S411: Based on the optimized location selection decision configuration, collect geographic information system and grid data, including terrain, climate, and grid capacity factors, to obtain the geographic and grid data set; Obtain the topographic data of the geographic information system through satellite remote sensing data and ground measurement equipment. After processing, the data can accurately display the undulation of the terrain, slope, and distribution of natural obstacles. Further obtain climate data, screen out areas with a high frequency of extreme weather and mark the risk points. At the same time, analyze the load capacity of the existing power grid, the distribution of power generation stations and substations, and transmission capacity through the power grid capacity report. After processing and cleaning the data, integrate it through data aggregation and marking technology to form a multi-dimensional geographic and power grid information dataset of terrain, climate, and power grid capacity. After completion, use the data for subsequent site selection evaluation and optimization processes.

[0051] S412: Based on the geographic and power grid dataset, determine the feasibility of the site selection location and potential difficulties in power grid access, and obtain the feasibility analysis result; Use spatial analysis technology to process the geographic conditions and power grid access data of each location. First, combine the topographic data with the power grid capacity distribution map through overlay analysis technology to find the obstacles caused by complex terrain areas to power grid wiring. Then, analyze the overlap degree between the high-risk areas marked in the climate data and the site selection to determine the environmental reliability of the site selection. At the same time, evaluate the power grid access load capacity of each candidate location. By comparing the existing capacity and estimated demand of the substation, calculate its remaining access capacity to judge whether it meets the needs of the new site selection point. In this process, combine the geographic and climate constraints and the power grid capacity calculation results to obtain the feasibility analysis result of the site selection location.

[0052] S413: Conduct a benefit analysis on the feasibility analysis result, using the formula:

[0053] Calculate the benefit index of the th site selection location, and continuously optimize the site selection decision in combination with geographic and power grid constraints to obtain the optimal site selection overview result; where, represents the value of the th evaluation parameter, representing the specific values considered in the site selection evaluation, such as power grid capacity, superiority of geographical location, etc., is the weight of the th evaluation parameter, indicating the importance of this parameter in the economic benefit evaluation, represents the direct cost of the th location, covering construction, operation and maintenance costs, etc., represents the potential risk adjustment value of the th location, is the total number of parameters, indicating the types of variables included in the calculation process; The value of (Consider three evaluation parameters: grid capacity, superiority of geographical location, and construction cost), (Value of grid capacity), (Value of superiority of geographical location), (Value of construction cost), (Weight of grid capacity), (Weight of geographical location), (Weight of construction cost), (Direct cost), (Risk adjustment value). Derivation process:

[0054]

[0055]

[0056] The result shows that the economic benefit index of the th location is 1.26, which means an economic benefit of 0.26 can be obtained. The ratio is higher than 1, indicating that the location selection of this location is feasible.

[0057] The above is only a preferred embodiment of the present invention, and it does not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A dynamic evaluation and optimization method for grid-side energy storage system site selection, characterized in that: The following steps are involved: Based on the grid-side demand, perform grid load data analysis, calculate the difference between peak load and minimum load, compare with historical load data, identify the fluctuation pattern of grid demand, and obtain a load fluctuation pattern diagram; Based on the load fluctuation pattern diagram, a long short-term memory network is used to perform a real-time simulation test according to real-time input power grid data, and by comparing the simulation results with the current power grid data, the prediction error of the energy storage system site selection is optimized to obtain a real-time simulation calibration model; Based on the real-time simulation calibration model, the site selection and scale of the grid-side energy storage system are evaluated, and multi-objective optimization is performed in combination with cost, reliability and environmental factors to analyze the energy storage demand of candidate locations, determine the optimal site selection and scale, and obtain the optimized site selection decision configuration; Based on the optimized site selection decision configuration, combined with the geographic information system, the mutual influence and constraints of geographic and power grid data are analyzed, and the linear programming algorithm is used to continuously optimize the site selection decision to obtain the most optimal site selection overview result.

2. The method for dynamic evaluation and optimization of grid-side energy storage system site selection according to claim 1, characterized in that: The calculation steps of the difference value are specifically as follows: Based on the grid demand, the grid load data within the target time period is collected, and the daily maximum and minimum loads are screened to obtain the daily load extreme value data set; Based on the daily load extreme value data set, the formula is used: ； Calculate the difference between peak load and minimum load , and obtain the difference value data set, where Indicates the highest load value monitored during the target time period. Indicates the lowest load value monitored during the same time period.

3. The method for dynamic evaluation and optimization of grid-side energy storage system site selection according to claim 1, characterized in that: The steps for obtaining the load fluctuation mode diagram are specifically as follows: Collect current and historical grid load data, adjust them into time series format, and obtain integrated load data; Based on the integrated load data, the fluctuation index with historical load data is calculated using the formula: ； Get in time Load fluctuation score ,in, Indicates at time The current load value, Represents the average value of historical load data, Represents the standard deviation of historical load data; Based on the load fluctuation score, abnormal time points beyond the normal fluctuation range are identified to obtain a load fluctuation pattern diagram.

4. The method for dynamic evaluation and optimization of grid-side energy storage system site selection according to claim 1, characterized in that: The execution steps of the real-time simulation test are specifically as follows: Based on the load fluctuation pattern diagram, real-time data of the power grid is collected and sorted, and whether the data format is consistent with the input requirements of the long short-term memory network is verified to obtain processed data results; Based on the processed data results, the data are input into a long short-term memory network for real-time simulation testing to simulate the behavior of the power grid and obtain simulation test output results.

5. The method for dynamic evaluation and optimization of grid-side energy storage system site selection according to claim 1, characterized in that: The steps for obtaining the real-time simulation calibration model are specifically as follows: Comparing the simulation results with the current grid data, the formula is used: ； Calculate the mean absolute error between the predicted value and the current value , and obtain the prediction error evaluation result, where Indicates Prediction data points, Indicates Current data points, Indicates the total number of data points; Based on the prediction error evaluation results, the error sources are analyzed, the network structure and learning parameters of the LSTM are adjusted, the power grid data is re-simulated and tested, the prediction error of the energy storage system site selection is optimized, and a real-time simulation calibration model is obtained.

6. The method for dynamic evaluation and optimization of grid-side energy storage system site selection according to claim 1, characterized in that: The specific steps for site selection and scale assessment are as follows: Based on the real-time simulation calibration model, real-time simulation is performed on potential energy storage sites to evaluate the geographical location, grid connection point and infrastructure support of each site, and obtain preliminary site performance evaluation data; Based on the preliminary performance evaluation data of the site, the formula is used: ； Calculate the Optimization index of sites , get the evaluation information of multi-site site selection, where Representative The power capacity of each site, Indicates The reliability index of each site, Indicates The construction and operation and maintenance costs of each site, yes The weight coefficient of yes The weight coefficient of yes The nonlinear adjustment coefficient of Based on the evaluation information of the multi-site site selection, the scale and configuration of the corresponding site are analyzed to obtain site selection and scale information.

7. The method for dynamic evaluation and optimization of grid-side energy storage system site selection according to claim 1, characterized in that: The steps for obtaining the optimized location decision configuration are specifically as follows: Combining the cost, reliability and environmental factors, a multi-objective optimization process is performed to score each candidate site, determine the energy storage demand and potential advantages of each site, and obtain a site selection score result; Based on the site selection scoring results, a location with a high score is determined as the most preferred site, and the scale of the site selection is determined according to the energy storage demand to obtain an optimized site selection decision configuration.

8. The method for dynamic evaluation and optimization of grid-side energy storage system site selection according to claim 1, characterized in that: The steps for obtaining the optimal address overview result are specifically as follows: Based on the optimized siting decision configuration, collecting geographic information system and power grid data, including topography, climate, and power grid capacity factors, to obtain a geographic and power grid data set; Based on the geographic and power grid data sets, determine the feasibility of the site selection and potential difficulties in grid access, and obtain feasibility analysis results; The feasibility analysis results are analyzed for benefits using the formula: ； Calculate the The benefit index of a site , combined with geographical and power grid constraints, continuously optimize site selection decisions and obtain an overview of the best site selection results; among them, Representative The value of the evaluation parameter, It is The weights of the evaluation parameters, Representative Direct costs at each location, Representative Potential risk-adjusted value for each location, is the total number of parameters.

Images