Charging energy storage adjustment processing method and system

Abstract

The invention discloses a charging energy storage adjustment processing method and system, through real-time monitoring and historical data prediction, in combination with an intelligent charging and discharging management strategy, the operation efficiency of a charging pile and the load balance capability of a power grid are improved, the accurate control of the system on the current condition is ensured through real-time data, and the real-time monitoring of the system is ensured. Historical data analysis enhances the accuracy of prediction, and dynamic charging and discharging management optimizes energy use, reduces power waste, and has positive influence on the stability of a power grid.

Description

Technical Field

[0001] The present invention relates to the field of charging energy storage of charging piles, and in particular to a charging energy storage adjustment processing method and system. Background Art

[0002] In traditional charging pile systems, the management of charging piles and the coordination of grid loads usually rely on fixed charging strategies, while ignoring the impact of real-time grid status and long-term grid changes on charging and discharging. In addition, traditional charging pile management systems often do not effectively monitor grid loads and charging pile status in real time. This is because the charging and discharging strategies of charging piles are usually based on static, preset conditions without considering actual grid load fluctuations. This may cause charging piles to consume excessive power during peak grid loads or fail to fully utilize power resources during low grid loads.

[0003] At the same time, although traditional re-installation systems store historical grid data and electricity price data, they often fail to make full use of these data for long-term predictions. As a result, the system may not be able to effectively predict seasonal changes and long-term trends in grid load, thus affecting the scheduling strategy of charging piles.

[0004] In addition, the charging and discharging management strategy of traditional charging piles is relatively simple, usually operating only based on fixed load thresholds and failing to make dynamic adjustments. Such a strategy may lead to waste of power resources or failure to effectively alleviate the pressure on the grid load. Summary of the invention

[0005] The purpose of the present invention is to provide a charging energy storage adjustment processing method and system to solve the above-mentioned technical problems pointed out in the prior art.

[0006] The present invention proposes a charging energy storage adjustment processing method, which includes the following steps:

[0007] Real-time monitoring to obtain current real-time status data; the real-time status data includes the charging pile status data of the current charging pile and the grid status data of the grid;

[0008] Collect and acquire historical data within a time period; perform data prediction operations based on the historical data and the real-time status data to obtain predicted power grid load data;

[0009] The historical data includes historical power grid status data and historical electricity price data;

[0010] The charging and discharging of the charging pile is managed and processed based on the predicted grid load data.

[0011] Preferably, the management and processing operations for charging and discharging of the charging pile based on the predicted grid load data include: when the predicted grid load data is lower than a preset minimum threshold value of the grid load data, controlling the charging pile to perform a charging operation; when the predicted grid load data is higher than a preset minimum threshold value of the grid load data, controlling the charging pile to perform a discharging operation.

[0012] Preferably, the data prediction operation is performed based on the historical data and the real-time status data to obtain the predicted power grid load data, including the following operation steps:

[0013] Perform optimization analysis based on the historical data to construct a time series model;

[0014] The real-time status data is input into the time series model, and the predicted power grid load data is output.

[0015] Preferably, the optimizing and analyzing the historical data to construct a time series model comprises the following steps:

[0016] Establish a time series diagram based on all the historical data; determine whether the time series diagram is stable based on the data fluctuation curve in the time series diagram; if so, use the historical data as the second historical data; if not, perform data sorting and processing operations on the historical data to obtain the second historical data;

[0017] Determine whether the new time series diagram formed by the second historical data is zero mean; if not, perform zero mean processing on the second historical data to obtain third historical data; if so, determine the second historical data as the third historical data;

[0018] The data sorting and processing operations include outlier detection and interpolation operations;

[0019] A second time series diagram is constructed based on the third historical data; an autocorrelation function ACF of each third historical data in a lag period β is calculated according to the second time series diagram; a partial autocorrelation function PACF of each third historical data in a second lag period β' is calculated according to the autocorrelation function ACF;

[0020] The third historical data includes a historical power grid load value α and a timestamp t corresponding to the historical power grid load value;

[0021] The model parameters of the time series model are obtained by performing calculation and analysis based on the autocorrelation function ACF and the partial autocorrelation function PACF;

[0022] The model parameters include autoregressive order a, moving average order b and difference order c;

[0023] A timing model is obtained based on the model parameters.

[0024] Preferably, the basis for judging whether the time series diagram is stable is to judge whether the absolute value of the difference between two adjacent historical data in the data fluctuation curve is less than or equal to a preset fluctuation judgment threshold.

[0025] Preferably, the autocorrelation function ACF is calculated as follows:

[0026]

[0027] In the formula, t is the timestamp; T is the length of the time series (or the number of timestamps); α t is the historical grid load value corresponding to timestamp t; β is the lag period; α t-β is the historical grid load value corresponding to the benchmark lag period with timestamp t; is the mean value of historical power grid load values;

[0028] The partial autocorrelation function PACF is calculated as follows:

[0029]

[0030] Where, β' is the second lag period; φ β'β is the autoregression coefficient.

[0031] Preferably, the step of calculating and analyzing the autocorrelation function ACF and the partial autocorrelation function PACF to obtain the model parameters of the time series model comprises the following steps:

[0032] Initialize the population to obtain multiple initial model parameters; at the same time, initialize the iteration parameters; the iteration parameters include the maximum iteration number threshold of the iteration counter, the model evaluation fitness threshold and the iteration counter; the iteration number of the iteration counter is initially 0;

[0033] Add one to the iteration number of the iteration counter to obtain the current iteration number; construct an initial time series model based on each of the initial model parameters; calculate and obtain the model evaluation fitness F(a, b, c) based on the initial time series model, the autocorrelation function ACF, and the partial autocorrelation function PACF;

[0034] Determine whether the model evaluation fitness F(a, b, c) is greater than or equal to a preset model evaluation fitness threshold; if so, output the current initial model parameters as target model parameters; if not, determine whether the current number of iterations is greater than or equal to a preset maximum number of iterations threshold of the iteration counter; if so, output the initial model parameters corresponding to the largest model evaluation fitness F(a, b, c) among the model evaluation fitnesses F(a, b, c) as target model parameters; if not, sort the model evaluation fitnesses F(a, b, c) from large to small to obtain a fitness sequence set;

[0035] Select the initial model parameters corresponding to the first n model evaluation fitness F(a, b, c) in the fitness sequence set as the parent model parameters; perform crossover mutation based on the parent model parameters to obtain multiple second model parameters; use the second model parameters as the initial model parameters, and return to the above operation until the target model parameters are output.

[0036] Preferably, the initial model parameters represent a combination of model parameters (a, b, c).

[0037] Preferably, the model evaluation fitness F(a, b, c) is calculated as follows:

[0038] F(a,b,c)=AIC(a,b,c)+λ·MSE ACF +μ·MSE PACF ;

[0039] Where AIC(a,b,c) is the prediction error of the initial time series model; MSE ACF MSE is the mean square error between the predicted autocorrelation function and the actual autocorrelation function of the initial time series model; PACF is the mean square error between the predicted partial autocorrelation function and the actual partial autocorrelation function of the initial time series model; λ and μ are weight coefficients.

[0040] Accordingly, the present invention also provides a charging energy storage adjustment processing system, including a real-time status monitoring module, a prediction module and a charging and discharging management module;

[0041] The real-time status monitoring module is used to monitor and obtain the current real-time status data in real time; the real-time status data includes the charging pile status data of the current charging pile and the grid status data of the grid;

[0042] The prediction module is used to collect historical data within a time period; perform data prediction operations based on the historical data and the real-time status data to obtain predicted power grid load data;

[0043] The historical data includes historical power grid status data and historical electricity price data;

[0044] The charging and discharging management module is used to manage and process the charging and discharging of the charging pile based on the predicted grid load data.

[0045] Compared with the prior art, the embodiments of the present invention have at least the following technical advantages:

[0046] By analyzing the above-mentioned charging energy storage adjustment processing method and system provided by the present invention, it can be known that in specific applications, the current system status can be fully understood through the status data obtained through real-time monitoring. The real-time status data can identify possible system problems or grid load anomalies, provide basic data for decision-making, and help maintain the normal operation of the charging pile; further, through the collection of historical data over a one-year period, the long-term load changes and electricity price trends of the grid can be fully understood, providing rich background information for the prediction of grid load data, making the predicted grid load data more accurate and reliable, which not only helps to predict the future load of the grid, but also can consider seasonal changes and long-term trends in the prediction; finally, charging and discharging management based on the predicted grid load data can optimize the use of the energy storage system, and charging operations can be performed when the grid load is lower than the preset minimum threshold, and excess power can be absorbed when the grid load is low to reduce power waste. Discharging operations can be performed when the grid load is higher than the minimum threshold, and additional power can be provided to the grid to relieve grid pressure, which not only helps to balance the grid load, but also improves the energy utilization efficiency and economic benefits of the charging pile;

[0047] The technical solution adopted in the present invention improves the operating efficiency of charging piles and the load balancing ability of the power grid through real-time monitoring and historical data prediction, combined with intelligent charging and discharging management strategies. Real-time data ensures the system's accurate control of the current situation, historical data analysis enhances the accuracy of the prediction, and dynamic charging and discharging management optimizes energy use, reduces electricity waste, and has a positive impact on the stability of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A schematic diagram of the main flow of a charging energy storage adjustment processing method according to the first embodiment of the present invention;

[0049] Figure 2 A schematic diagram of the operation steps for obtaining predicted grid load data in a charging energy storage adjustment processing method according to the first embodiment of the present invention;

[0050] Figure 3 A schematic diagram of the operation flow of constructing a timing model in a charging energy storage adjustment processing method according to the first embodiment of the present invention;

[0051] Figure 4 A schematic diagram of interpolation simulation of a power grid load data curve in a charging energy storage adjustment processing method according to the first embodiment of the present invention;

[0052] Figure 5 It is a schematic diagram of the operation flow of analyzing and obtaining the model parameters of the timing model in a charging energy storage adjustment processing method according to the first embodiment of the present invention;

[0053] Figure 6 This is a schematic diagram of the overall architecture of a charging energy storage adjustment processing system according to the second embodiment of the present invention.

[0054] Reference numerals: real-time status monitoring module 10 , prediction module 20 , charge and discharge management module 30 , time series model building module 21 , prediction output module 22 . DETAILED DESCRIPTION

[0055] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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.

[0056] The present invention is further described in detail below through specific embodiments in conjunction with the accompanying drawings.

[0057] Embodiment 1

[0058] like Figure 1 As shown, the first embodiment of the present invention provides a charging energy storage adjustment processing method, including the following operation steps:

[0059] Step S10: Real-time monitoring to obtain current real-time status data; the real-time status data includes the charging pile status data of the current charging pile and the grid status data of the grid;

[0060] Step S20: collecting historical data within a time period; performing data prediction operations based on the historical data and the real-time status data to obtain predicted power grid load data;

[0061] The historical data includes historical power grid status data and historical electricity price data;

[0062] The above time period generally refers to one year; collecting historical data in a one-year period can more comprehensively predict the predicted grid load data of the current charging pile on the grid in the subsequent prediction process;

[0063] Step S30: performing management and processing operations on the charging and discharging of the charging pile based on the predicted grid load data;

[0064] Specifically, the management and processing operation of charging and discharging the charging pile based on the predicted grid load data includes: when the predicted grid load data is lower than the preset grid load data minimum threshold, controlling the energy storage system of the charging pile (by the grid) to perform charging operation; when the predicted grid load data is higher than the preset grid load data minimum threshold, controlling the energy storage system of the charging pile to perform discharging operation (to the grid);

[0065] It should be noted that the above-mentioned implementation of this application obtains a comprehensive understanding of the current system status through real-time monitoring of status data (including charging pile status and grid status). Real-time status data can identify possible system problems or grid load anomalies, provide basic data for decision-making, and help maintain the normal operation of charging piles; further, through the collection of historical data (grid status data and electricity price data) over a one-year period, a comprehensive understanding of the long-term load changes and electricity price trends of the grid can be obtained, providing rich background information for the prediction of grid load data, making the predicted grid load data more accurate and reliable, which not only helps to predict the future load of the grid, but also considers seasonal changes and long-term trends in the prediction; finally, charging and discharging management based on predicted grid load data can optimize the use of the energy storage system, perform charging operations when the grid load is lower than the preset minimum threshold, absorb excess power when the grid load is low, reduce power waste, and perform discharging operations when the grid load is higher than the minimum threshold, which can provide additional power to the grid and relieve grid pressure, which not only helps to balance the grid load, but also improves the energy utilization efficiency and economic benefits of charging piles;

[0066] The embodiments of the present application combine real-time monitoring and historical data prediction with intelligent charging and discharging management strategies to improve the operating efficiency of charging piles and the load balancing capability of the power grid. Real-time data ensures that the system has accurate control over the current situation, historical data analysis enhances the accuracy of the prediction, and dynamic charging and discharging management optimizes energy use, reduces electricity waste, and has a positive impact on the stability of the power grid.

[0067] Specifically, Figure 2 As shown, in step S20, a data prediction operation is performed based on the historical data and the real-time status data to obtain predicted power grid load data, including the following operation steps:

[0068] Step S21: Optimizing and analyzing the historical data to construct a time series model;

[0069] Step S22: inputting the real-time status data into the time series model, and outputting the predicted power grid load data;

[0070] It should be noted that the above-mentioned embodiment of the present application uses historical data (including historical grid status and electricity price data) for optimization analysis to construct a time series model. Through optimization analysis, the model can capture the historical trends and patterns of grid load, thereby improving the prediction accuracy. However, the key to the embodiment of the present application is to create a model that can process time series data so that it can effectively reflect the change law of grid load over time; further, by inputting real-time status data (charging pile status and grid status) into the constructed time series model, the model processes these data and outputs predicted grid load data; in this way, real-time data is combined with historical data to generate an accurate prediction of future grid load, thereby providing a basis for charging and discharging management of charging piles;

[0071] The above-mentioned embodiments of the present application realize accurate prediction of grid load, construct a time series model through optimization analysis, ensure the reasonable description of grid load changes by the prediction model, and further combine real-time data with the time series model to provide instant load prediction; the implementation of the embodiments of the present application improves the accuracy and real-time performance of the prediction, helps to optimize the charging and discharging operations of the charging piles, and thereby improves the energy utilization efficiency of the overall system and the stability of the grid.

[0072] During the specific implementation of the above-mentioned embodiment of the present application, technical personnel have discovered that the time series model established by the operation and maintenance historical data will have fluctuations on the trend curve of the data because the historical data is sample data of a non-stationary time series. Therefore, the predicted power grid load data obtained by predicting the time series model will result in fluctuating errors in the results. Therefore, in the subsequent prediction process, it is necessary to adjust and optimize the time series model to obtain a more stable time series model that is more in line with the actual situation of the historical data, so as to obtain more accurate predicted power grid load data.

[0073] Specifically, Figure 3 As shown, in step S22, optimizing and analyzing the historical data to construct a time series model includes the following steps:

[0074] Step S221: Establish a time series diagram based on all the historical data; determine whether the time series diagram is stable based on the data fluctuation curve in the time series diagram; if not (if yes, directly use the historical data as the second historical data to perform a zero mean test operation), perform a data sorting and processing operation on the historical data to obtain the second historical data; determine whether the new time series diagram formed by the second historical data is zero mean for verification; if not (if yes, determine the second historical data as the third historical data and perform the next step), perform zero mean processing on the second historical data to obtain the third historical data;

[0075] The data sorting and processing operations include outlier detection and interpolation operations;

[0076] The basis for judging whether the time series graph is stable is to judge whether the absolute value of the difference between two adjacent historical data in the data fluctuation curve is less than or equal to a preset fluctuation judgment threshold;

[0077] In the above-mentioned embodiment of the present application, if the power grid load data suddenly increases abnormally at a specific time, it may be an abnormality caused by a data entry error or a special event. Therefore, it is necessary to perform data sorting and processing operations on the historical data to eliminate the influence of the abnormal data on the subsequent prediction results and provide prediction accuracy;

[0078] The above-mentioned zero mean refers to the probability distribution of a random variable with an expected value of 0; the characteristics of this distribution are: its probability density function or probability distribution function is non-negative in all possible values, and the overall probability is 1; in statistics and machine learning, zero mean is a common assumption. For example, under the assumption of normal distribution, the mean of the data is usually assumed to be 0; zero mean can simplify the parameterization and calculation process of the model, and can also improve the generalization ability of the model.

[0079] Step S222: construct a second time series diagram based on the third historical data; calculate the autocorrelation function ACF of each third historical data in the lag period β according to the second time series diagram; calculate the partial autocorrelation function PACF of each third historical data in the second lag period β' according to the autocorrelation function ACF;

[0080] The third historical data includes a historical power grid load value α and a timestamp t corresponding to the historical power grid load value;

[0081] The autocorrelation function ACF is calculated as follows:

[0082]

[0083] In the formula, t is the timestamp; T is the length of the time series (or the number of timestamps); α t is the historical grid load value corresponding to timestamp t; β is the lag period; α t-β is the historical grid load value corresponding to the benchmark lag period with timestamp t; is the mean value of historical power grid load values;

[0084] The partial autocorrelation function PACF is calculated as follows:

[0085]

[0086] Where, β' is the second lag period; φ β'β is the autoregression coefficient;

[0087] Step S223: Calculating and analyzing the autocorrelation function ACF and the partial autocorrelation function PACF to obtain the model parameters of the time series model;

[0088] The model parameters include autoregressive order a, moving average order b and difference order c;

[0089] Step S224: Obtain a timing model based on the model parameters.

[0090] It should be noted that the above-mentioned embodiment of the present application first determines whether the data is stable by establishing a time series diagram and analyzing data fluctuations. If the data is not stable, data sorting processing (such as outlier detection and interpolation) is performed on it. For example: Figure 4 As shown, in the embodiment of the present application, the temperature gradually rises from January to June. Therefore, the use of household appliances, such as fans and air conditioners, will cause the grid load to increase. The solid line part is the actually collected grid load data. It can be seen that in June, due to data entry errors or special events, the grid load data curve shows obvious fluctuations, which is inconsistent with the actual situation; therefore, according to the data sorting and processing operation (i.e., interpolation operation) in the embodiment of the present application, the grid load data of the dotted line part in June is obtained, so that the data curve is stable;

[0091] Then, it is further verified whether the processed data is zero mean, and the data is further adjusted to ensure that its mean is zero; thereby cleaning and normalizing the data, eliminating the impact of outliers on model training, and ensuring that the data is suitable for subsequent modeling; further using the third historical data to construct a time series diagram, and calculating the autocorrelation function (ACF) and partial autocorrelation function (PACF), the autocorrelation function ACF helps to determine whether there is a lag effect and periodicity in the data, providing a basis for the selection of model parameters, and the partial autocorrelation function PACF further refines the autocorrelation structure of the data, helping to identify whether it is necessary to add autoregressive or moving average components, and can quickly and accurately analyze the time dependence of the data, thereby identifying the periodicity and autocorrelation structure in the data, providing a basis for the parameter selection of the time series model; then based on the ACF and PACF obtained in step S222, the parameters of the time series model are calculated, including the autoregressive order (a), the moving average order (b) and the number of differences (c); finally, the model parameters determined in step S223 are used to construct the final time series model; this model will be used to predict the power grid load;

[0092] The implementation of the above-mentioned embodiments of the present application can ensure that the data preprocessing, model construction and parameter optimization processes are scientific and accurate, effectively improve the accuracy and reliability of the prediction model, reduce the interference of abnormal data on the prediction results, thereby improving the accuracy of power grid load prediction, and ultimately optimizing power grid management and operation efficiency.

[0093] Explanation: The third historical data includes a historical power grid load value α and a timestamp t corresponding to the historical power grid load value;

[0094] However, the autocorrelation function ACF is designed to be related to the timestamp, the length of the time series, the historical power grid load value corresponding to the timestamp t, the hysteresis period, the mean of the historical power grid load value, etc.; α is the historical power grid load value corresponding to the timestamp t; β is the hysteresis period; is the mean value of the historical power grid load values. The following describes the process of obtaining the model parameters of the time series model by calculating and analyzing the autocorrelation function ACF and the partial autocorrelation function PACF.

[0095] Specifically, Figure 5 As shown, in step S223, the model parameters of the time series model are obtained by calculation and analysis based on the autocorrelation function ACF and the partial autocorrelation function PACF, including the following operation steps:

[0096] Step S2231: Initialize the population to obtain multiple initial model parameters; at the same time, initialize the iteration parameters; the iteration parameters include the maximum iteration number threshold of the iteration counter, the model evaluation fitness threshold and the iteration counter; the iteration number of the iteration counter is initially 0;

[0097] Each initial model parameter obtained by the above-mentioned initialization population is an individual in the population, and each individual represents a combination of model parameters (a, b, c);

[0098] Step S2232: adding one to the iteration number of the iteration counter to obtain the current iteration number; constructing an initial time series model based on each of the initial model parameters; and calculating and obtaining the model evaluation fitness F(a, b, c) based on the initial time series model and the autocorrelation function ACF and the partial autocorrelation function PACF;

[0099] The calculation method of the model evaluation fitness F(a, b, c) is:

[0100] F(a,b,c)=AIC(a,b,c)+λ·MSE ACF +μ·MSE PACF ;

[0101] Where AIC(a,b,c) is the prediction error of the initial time series model; MSE ACF MSE is the mean square error between the predicted autocorrelation function and the actual autocorrelation function of the initial time series model; PACF is the mean square error between the predicted partial autocorrelation function and the actual partial autocorrelation function of the initial time series model; λ and μ are weight coefficients;

[0102] Step S2233: Determine whether the model evaluation fitness F(a, b, c) is greater than or equal to a preset model evaluation fitness threshold; if so, output the current initial model parameters as target model parameters; if not, determine whether the current number of iterations is greater than or equal to a preset maximum number of iterations threshold of the iteration counter; if so, output the initial model parameters corresponding to the largest model evaluation fitness F(a, b, c) among the model evaluation fitnesses F(a, b, c) as target model parameters; if not, sort the model evaluation fitnesses F(a, b, c) from large to small to obtain a fitness sequence set;

[0103] Step S2234: select the initial model parameters corresponding to the first n model evaluation fitness F(a, b, c) in the fitness sequence set as the parent model parameters; perform crossover mutation based on the parent model parameters to obtain multiple second model parameters; use the second model parameters as the initial model parameters, and return to the above operation until the target model parameters are output.

[0104] It should be noted that the above-mentioned embodiment of the present application first generates multiple initial model parameter combinations (a, b, c), each combination represents an individual, to construct an initial time series model. By initializing multiple model parameter combinations, multiple solution candidates are provided, increasing the chance of finding the best model parameters; the control of the optimization process is ensured by setting the iterative parameters to avoid the problems of excessive iteration and non-convergence; the model parameters are further updated and evaluated by each iteration, the model performance is gradually optimized, and a comprehensive evaluation is performed in combination with AIC, ACF and PACF to improve the prediction accuracy and stability of the model; by checking whether the fitness of the current model meets the preset threshold, if it meets, the current parameters are output as the target model, if the fitness threshold is not met and the maximum number of iterations is not reached, the iteration continues, otherwise the optimal model parameters are output; according to the fitness threshold or the maximum number of iterations, the optimization process is stopped in time to avoid unnecessary calculations, and the model parameters with the highest fitness are selected as the final result to ensure the quality of the model. When the final target model parameters cannot be obtained by summarizing the iterative process, new model parameter combinations are explored through crossover and mutation operations to increase the possibility of finding a better solution, continuously update and optimize the model parameters, and improve the prediction performance of the overall model;

[0105] The above-mentioned embodiments of the present application systematically optimize model parameters through iteration and evaluation, improve the prediction ability of the time series model, utilize crossover mutation strategies to improve the diversity of solutions and optimization speed, find the optimal model parameters, and ensure the stability and efficiency of the optimization process by setting fitness and iterative control parameters.

[0106] Embodiment 2

[0107] like Figure 6As shown, the second embodiment of the present invention provides a charging energy storage adjustment processing system, including a real-time status monitoring module 10, a prediction module 20 and a charging and discharging management module 30;

[0108] The real-time status monitoring module 10 is used to monitor and obtain the current real-time status data in real time; the real-time status data includes the charging pile status data of the current charging pile and the grid status data of the grid;

[0109] The prediction module 20 is used to collect historical data within a time period; perform data prediction operations based on the historical data and the real-time status data to obtain predicted power grid load data;

[0110] The historical data includes historical power grid status data and historical electricity price data;

[0111] The charging and discharging management module 30 is used to manage and process the charging and discharging of the charging pile based on the predicted grid load data;

[0112] Preferably, the charging and discharging management module 30 is specifically used to control the energy storage system of the charging pile (by the power grid) to perform charging operations when the predicted power grid load data is lower than the preset minimum threshold value of the power grid load data; and to control the energy storage system of the charging pile to perform discharging operations (to the power grid) when the predicted power grid load data is higher than the preset minimum threshold value of the power grid load data.

[0113] Preferably, the prediction module 20 includes a time series model building module 21 and a prediction output module 22;

[0114] The time series model building module 21 is used to optimize and analyze the historical data to build a time series model;

[0115] The prediction output module 22 is used to input the real-time status data into the time series model and output predicted power grid load data.

[0116] Specifically, the time series model building module 21 is specifically used to establish a time series diagram based on all the historical data; determine whether the time series diagram is stable based on the data fluctuation curve in the time series diagram; if not (if yes, directly use the historical data as the second historical data to perform a zero mean test operation), perform a data sorting and processing operation on the historical data to obtain the second historical data; determine whether the new time series diagram formed by the second historical data is zero mean; if not (if yes, determine the second historical data as the third historical data and perform the next step), perform zero mean processing on the second historical data to obtain the third historical data;

[0117] The data sorting and processing operations include outlier detection and interpolation operations;

[0118] The basis for judging whether the time series graph is stable is to judge whether the absolute value of the difference between two adjacent historical data in the data fluctuation curve is less than or equal to a preset fluctuation judgment threshold;

[0119] A second time series diagram is constructed based on the third historical data; an autocorrelation function ACF of each third historical data in a lag period β is calculated according to the second time series diagram; a partial autocorrelation function PACF of each third historical data in a second lag period β' is calculated according to the autocorrelation function ACF;

[0120] The third historical data includes a historical power grid load value α and a timestamp t corresponding to the historical power grid load value;

[0121] The autocorrelation function ACF is calculated as follows:

[0122]

[0123] In the formula, t is the timestamp; T is the length of the time series (or the number of timestamps); α t is the historical grid load value corresponding to timestamp t; β is the lag period; α t-β is the historical grid load value corresponding to the benchmark lag period with timestamp t; is the mean value of historical power grid load values;

[0124] The partial autocorrelation function PACF is calculated as follows:

[0125]

[0126] Where, β' is the second lag period; φ β'β is the autoregression coefficient;

[0127] The model parameters of the time series model are obtained by performing calculation and analysis based on the autocorrelation function ACF and the partial autocorrelation function PACF;

[0128] The model parameters include autoregressive order a, moving average order b and difference order c;

[0129] A timing model is obtained based on the model parameters.

[0130] In the specific implementation process, the time series model building module 21 is also used to initialize the population and obtain multiple initial model parameters; at the same time, initialize the iteration parameters; the iteration parameters include the maximum iteration number threshold of the iteration counter, the model evaluation fitness threshold and the iteration counter; the iteration number of the iteration counter is initially 0;

[0131] Each initial model parameter obtained by the above-mentioned initialization population is an individual in the population, and each individual represents a combination of model parameters (a, b, c);

[0132] Add one to the iteration number of the iteration counter to obtain the current iteration number; construct an initial time series model based on each of the initial model parameters; calculate and obtain the model evaluation fitness F(a, b, c) based on the initial time series model, the autocorrelation function ACF, and the partial autocorrelation function PACF;

[0133] The calculation method of the model evaluation fitness F(a, b, c) is:

[0134] F(a,b,c)=AIC(a,b,c)+λ·MSE ACF +μ·MSE PACF ;

[0135] Where AIC(a,b,c) is the prediction error of the initial time series model; MSE ACF MSE is the mean square error between the predicted autocorrelation function and the actual autocorrelation function of the initial time series model; PACF is the mean square error between the predicted partial autocorrelation function and the actual partial autocorrelation function of the initial time series model; λ and μ are weight coefficients;

[0136] Determine whether the model evaluation fitness F(a, b, c) is greater than or equal to a preset model evaluation fitness threshold; if so, output the current initial model parameters as target model parameters; if not, determine whether the current number of iterations is greater than or equal to a preset maximum number of iterations threshold of the iteration counter; if so, output the initial model parameters corresponding to the largest model evaluation fitness F(a, b, c) among the model evaluation fitnesses F(a, b, c) as target model parameters; if not, sort the model evaluation fitnesses F(a, b, c) from large to small to obtain a fitness sequence set;

[0137] Select the initial model parameters corresponding to the first n model evaluation fitness F(a, b, c) in the fitness sequence set as the parent model parameters; perform crossover mutation based on the parent model parameters to obtain multiple second model parameters; use the second model parameters as the initial model parameters, and return to the above operation until the target model parameters are output.

[0138] In summary, the charging energy storage adjustment processing method and system proposed in the example of the present invention can fully understand the current system status through the status data (including the charging pile status and the power grid status) obtained through real-time monitoring. The real-time status data can identify possible system problems or power grid load anomalies, provide basic data for decision-making, and help maintain the normal operation of the charging pile; further, through the collection of historical data (power grid status data and electricity price data) over a one-year period, the long-term load changes and electricity price trends of the power grid can be fully understood, providing rich background information for predicting the power grid load data, making the predicted power grid load data more accurate and reliable, which not only helps to predict the future load of the power grid, but also can consider seasonal changes and long-term trends in the prediction; in specific operations, a time series model is constructed through optimization analysis to ensure the reasonable description of the power grid load changes by the prediction model, and further combine the real-time data with the time series model to provide instant load prediction;

[0139] During the specific implementation process, the accuracy and reliability of the prediction model were effectively improved through data preprocessing, model construction and parameter optimization, and the interference of abnormal data on the prediction results was reduced, thereby improving the accuracy of power grid load prediction and ultimately optimizing power grid management and operation efficiency;

[0140] Furthermore, through iteration and evaluation, the model parameters are systematically optimized to improve the prediction ability of the time series model. The crossover mutation strategy is used to improve the diversity of solutions and the optimization speed, and the optimal model parameters are found. By setting the fitness and iterative control parameters, the stability and efficiency of the optimization process are guaranteed.

[0141] Ultimately, charging and discharging management based on predicted grid load data can optimize the use of energy storage systems. Charging operations can be performed when the grid load is below the preset minimum threshold, absorbing excess power when the grid load is low and reducing power waste. Discharging operations can be performed when the grid load is above the minimum threshold, providing additional power to the grid and relieving grid pressure. This not only helps balance the grid load, but also improves the energy efficiency and economic benefits of charging piles.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. A person skilled in the art may modify the technical solutions described in the above embodiments, or replace part or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A charging energy storage adjustment processing method, characterized in that: The steps are as follows: Real-time monitoring to obtain current real-time status data; the real-time status data includes the charging pile status data of the current charging pile and the grid status data of the grid; Collect and acquire historical data within a time period; perform data prediction operations based on the historical data and the real-time status data to obtain predicted power grid load data; The historical data includes historical power grid status data and historical electricity price data; The charging and discharging of the charging pile is managed and processed based on the predicted grid load data.

2. A charging energy storage adjustment processing method according to claim 1, characterized in that: The management and processing operations for charging and discharging of the charging pile based on the predicted grid load data include: when the predicted grid load data is lower than the preset minimum threshold of the grid load data, controlling the charging pile to perform a charging operation; when the predicted grid load data is higher than the preset minimum threshold of the grid load data, controlling the charging pile to perform a discharging operation.

3. A charging energy storage adjustment processing method according to claim 2, characterized in that: The performing of data prediction operation based on the historical data and the real-time status data to obtain predicted power grid load data comprises the following operation steps: Perform optimization analysis based on the historical data to construct a time series model; The real-time status data is input into the time series model, and the predicted power grid load data is output.

4. A charging energy storage adjustment processing method according to claim 3, characterized in that: The optimizing analysis based on the historical data to construct a time series model comprises the following steps: Establish a time series diagram based on all the historical data; determine whether the time series diagram is stable based on the data fluctuation curve in the time series diagram; if so, use the historical data as the second historical data; if not, perform data sorting and processing operations on the historical data to obtain the second historical data; Determine whether the new time series diagram formed by the second historical data is zero mean; if not, perform zero mean processing on the second historical data to obtain third historical data; if so, determine the second historical data as the third historical data; The data sorting and processing operations include outlier detection and interpolation operations; A second time series diagram is constructed based on the third historical data; an autocorrelation function ACF of each third historical data in a lag period β is calculated according to the second time series diagram; a partial autocorrelation function PACF of each third historical data in a second lag period β' is calculated according to the autocorrelation function ACF; The third historical data includes a historical power grid load value α and a timestamp t corresponding to the historical power grid load value; The model parameters of the time series model are obtained by performing calculation and analysis based on the autocorrelation function ACF and the partial autocorrelation function PACF; The model parameters include autoregressive order a, moving average order b and difference order c; A timing model is obtained based on the model parameters.

5. A charging energy storage adjustment processing method according to claim 4, characterized in that: The basis for judging whether the time series diagram is stable is to judge whether the absolute value of the difference between two adjacent historical data in the data fluctuation curve is less than or equal to a preset fluctuation judgment threshold.

6. A charging energy storage adjustment processing method according to claim 5, characterized in that: The autocorrelation function ACF is calculated as follows: Where t is the timestamp; T is the length of the time series (or the number of timestamps); α t is the historical grid load value corresponding to timestamp t; β is the lag period; α t-β is the historical grid load value corresponding to the benchmark lag period with timestamp t; is the mean value of historical power grid load values; The partial autocorrelation function PACF is calculated as follows: Where, β' is the second lag period; φ β'β is the autoregression coefficient.

7. A charging energy storage adjustment processing method according to claim 6, characterized in that: The step of calculating and analyzing the autocorrelation function ACF and the partial autocorrelation function PACF to obtain the model parameters of the time series model includes the following steps: Initialize the population to obtain multiple initial model parameters; at the same time, initialize the iteration parameters; the iteration parameters include the maximum iteration number threshold of the iteration counter, the model evaluation fitness threshold and the iteration counter; the iteration number of the iteration counter is initially 0; Add one to the iteration number of the iteration counter to obtain the current iteration number; construct an initial time series model based on each of the initial model parameters; calculate and obtain the model evaluation fitness F(a, b, c) based on the initial time series model, the autocorrelation function ACF, and the partial autocorrelation function PACF; Determine whether the model evaluation fitness F(a, b, c) is greater than or equal to a preset model evaluation fitness threshold; if so, output the current initial model parameters as target model parameters; if not, determine whether the current number of iterations is greater than or equal to a preset maximum number of iterations threshold of the iteration counter; if so, output the initial model parameters corresponding to the largest model evaluation fitness F(a, b, c) among the model evaluation fitness F(a, b, c) as the target model parameters; if not, sort the model evaluation fitness F(a, b, c) from large to small to obtain a fitness sequence set; Select the initial model parameters corresponding to the first n model evaluation fitness F(a, b, c) in the fitness sequence set as the parent model parameters; perform crossover mutation based on the parent model parameters to obtain multiple second model parameters; use the second model parameters as the initial model parameters, and return to the above operation until the target model parameters are output.

8. A charging energy storage adjustment processing method according to claim 7, characterized in that: The initial model parameters represent a combination of model parameters (a, b, c).

9. A charging energy storage adjustment processing method according to claim 8, characterized in that: The calculation method of the model evaluation fitness F(a, b, c) is: F(a,b,c)=AIC(a,b,c)+λ·MSE ACF +μ·MSE PACF ; Where AIC(a,b,c) is the prediction error of the initial time series model; MSE ACF MSE is the mean square error between the predicted autocorrelation function and the actual autocorrelation function of the initial time series model; PACF is the mean square error between the predicted partial autocorrelation function and the actual partial autocorrelation function of the initial time series model; λ and μ are weight coefficients.

10. A charging energy storage adjustment processing system, characterized in that: A charging energy storage adjustment processing method as claimed in any one of claims 1 to 9 is utilized, comprising a real-time status monitoring module, a prediction module and a charging and discharging management module; The real-time status monitoring module is used to monitor and obtain the current real-time status data in real time; the real-time status data includes the charging pile status data of the current charging pile and the grid status data of the grid; The prediction module is used to collect historical data within a time period; perform data prediction operations based on the historical data and the real-time status data to obtain predicted power grid load data; The historical data includes historical power grid status data and historical electricity price data; The charging and discharging management module is used to manage and process the charging and discharging of the charging pile based on the predicted grid load data.

Images