Energy storage equipment charging and discharging control system based on artificial intelligence

Abstract

The invention discloses an energy storage equipment charging and discharging control system based on artificial intelligence. The system comprises a data acquisition module, a data preprocessing module, a power demand prediction module, an energy storage equipment charging and discharging prediction module, an equipment charging and discharging strategy optimization module and an energy storage equipment charging and discharging intelligent control module. The invention relates to the technical field of data processing of energy storage equipment, in particular to an energy storage equipment charge and discharge control system based on artificial intelligence, which innovatively proposes that the intelligent control of charge and discharge of the energy storage equipment is realized by predicting the power demand and the state of the energy storage equipment in real time and combining an optimization algorithm to adjust a charge and discharge strategy; the long-term behavior features and the short-term dynamic charging and discharging features are combined, and feature fusion is performed by using a dynamic weighting mechanism, so that the accuracy of energy storage equipment charging and discharging state prediction is improved; by improving the particle optimization algorithm for obtaining the optimal charging and discharging strategy of the energy storage equipment, the global optimal solution search capability is improved, and the optimal charging and discharging strategy of the energy storage equipment is obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage device data processing, and in particular to an energy storage device charging and discharging control system based on artificial intelligence. Background Art

[0002] With the continuous development of renewable energy applications, energy production and consumption are highly volatile and intermittent, which brings challenges to the stability and reliability of the power system. Therefore, energy storage technology is increasingly widely used in power systems. Energy storage equipment can store energy when power demand is low and release energy when demand is high, thereby improving the stability and reliability of the system. To meet this challenge, an energy storage equipment charging and discharging control system based on artificial intelligence has emerged. Through the collection and intelligent analysis of real-time data, it can dynamically optimize the charging and discharging process of energy storage equipment, improve energy utilization efficiency, reduce energy waste, extend the service life of energy storage equipment, and realize intelligent control of energy storage equipment charging and discharging, providing strong support for the sustainable development of the power system; However, there are technical problems in the traditional energy storage equipment charging and discharging control system, such as low energy storage equipment life, low charging and discharging efficiency, and uncontrollable charging and discharging process; the existing prediction model for energy storage equipment charging and discharging has problems such as insufficient feature extraction, insufficient time dependency modeling, and limited ability to handle nonlinear features, which leads to inaccurate prediction of the charging and discharging state of energy storage equipment; the traditional equipment charging and discharging strategy optimization algorithm has weak ability to obtain the global optimal solution, which leads to the technical problem that the final charging and discharging strategy is not suitable. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides an energy storage device charging and discharging control system based on artificial intelligence. In view of the technical problems of low energy storage device life, low charging and discharging efficiency and uncontrollable charging and discharging process in traditional energy storage device charging and discharging control systems, this solution innovatively proposes to predict power demand and energy storage device status in real time, and dynamically adjust the charging and discharging strategy in combination with optimization algorithms. It can flexibly select the best charging and discharging time for energy storage devices, ensure that the energy storage devices are charged and discharged in the optimal time period, maximize the charging and discharging efficiency, reduce energy loss, effectively avoid excessive charging and discharging, extend the service life of energy storage devices and reduce the battery attenuation rate, improve the stability and responsiveness of the charging and discharging control system, solve the problems of power system overload and insufficient power of energy storage devices, and improve the operating efficiency of energy storage devices, thereby realizing intelligent control of energy storage devices in the region; in view of the insufficient feature extraction and insufficient time dependency modeling in the existing prediction models for charging and discharging of energy storage devices As well as the problem of limited ability to handle nonlinear characteristics, which leads to inaccurate prediction of the charging and discharging status of energy storage equipment. This solution innovatively proposes a method that combines long-term behavioral feature extraction with short-term dynamic charging and discharging feature modeling. By using a dynamic weighting mechanism and feature fusion strategy, it solves the problem of insufficient capture of long-term trends and short-term fluctuation characteristics by traditional models, and improves the adaptability and flexibility of the model to complex energy storage equipment charging and discharging prediction scenarios, thereby improving the accuracy of the output results of energy storage equipment charging and discharging predictions, and providing more data support for the intelligent control of energy storage equipment charging and discharging; in view of the technical problem that the traditional equipment charging and discharging strategy optimization algorithm has a weak ability to obtain the global optimal solution, which leads to the final charging and discharging strategy being unsuitable, this solution enhances the global optimal solution search and improves the flexibility of the algorithm by designing particle position constraints, fitness functions, nonlinear inertia weight strategies and eliminating inferior particles strategies, thereby obtaining the optimal charging and discharging strategy suitable for energy storage equipment and meeting the intelligent charging and discharging control of energy storage equipment.

[0004] The technical solution adopted by the present invention is as follows: an energy storage device charging and discharging control system based on artificial intelligence provided by the present invention includes a data acquisition module, a data preprocessing module, a power demand prediction module, an energy storage device charging and discharging prediction module, an equipment charging and discharging strategy optimization module and an energy storage device charging and discharging intelligent control module;

[0005] The data acquisition module acquires the original data of charge and discharge control of energy storage equipment by collecting data from the national power grid system;

[0006] The data preprocessing module performs data cleaning, time alignment, standardization processing and feature selection on the original data of the energy storage device charge and discharge control to obtain preliminary data of the energy storage device charge and discharge control;

[0007] The power demand forecasting module is used to forecast the future power demand in the region, and to build a power demand forecasting model, train the power demand forecasting model, realize real-time forecasting of power demand, and obtain the real-time power demand forecasting result of the region;

[0008] The energy storage device charge and discharge prediction module constructs an energy storage device charge and discharge prediction model, trains the energy storage device charge and discharge prediction model, realizes the real-time charge and discharge state prediction of the energy storage device, and obtains the real-time charge and discharge state prediction result of the energy storage device;

[0009] The device charging and discharging strategy optimization module obtains the optimal charging and discharging strategy of the energy storage device by improving the particle swarm optimization algorithm;

[0010] The energy storage device charge and discharge intelligent control module inputs data into the device charge and discharge strategy optimization module to obtain the optimal charge and discharge strategy of the energy storage device, and executes this strategy to realize intelligent charge and discharge control of the energy storage device.

[0011] Furthermore, the data acquisition module specifically acquires the original data of energy storage device charge and discharge control from the national power grid system by acquisition; the original data of energy storage device charge and discharge control includes historical power demand data, real-time power demand data, historical energy storage device data and real-time energy storage device data; the historical power demand data and real-time power demand data both include power load data, time characteristic data and weather data; the historical energy demand data also includes historical energy demand results; the historical energy storage device data and real-time energy storage device data both include device status data and device charge and discharge data.

[0012] Furthermore, the data preprocessing module specifically performs data cleaning, time alignment, standardization and feature selection on the original data of the energy storage device charge and discharge control to obtain preliminary data of the energy storage device charge and discharge control; the data cleaning is to process missing values, abnormal values ​​and duplicate values; the time alignment is to unify the time granularity of the data through time grouping and aggregation, unify the time granularity to the hourly level, and align the timestamps of data from different sources; the standardization is based on the maximum and minimum normalization method to standardize the data; the feature selection is to use the correlation analysis method to screen out the relevant features with the power demand forecast and the energy storage device charge and discharge forecast.

[0013] Furthermore, the power demand forecasting module is used to forecast future power demand in the region, help energy storage equipment to reasonably dispatch charging and discharging strategies, and avoid unnecessary high-cost operations, including power demand forecasting model construction, power demand forecasting model training, and regional real-time power demand forecasting; specifically, the following steps are included:

[0014] The construction of the power demand forecasting model includes the following steps:

[0015] The formula used for extracting the time series features of power demand is as follows:

[0016] ;

[0017] In the formula, Indicates the unit runs the function, represents the hidden state at the kth time step, represents the input data at the kth time step, Indicates that it is The hidden state of time steps, Indicates that it is The cell state at time steps,

[0018] The attention weight is calculated using the following formula:

[0019] ;

[0020] In the formula, represents the attention weight value of the kth time step, represents the attention weight matrix, represents the attention bias parameter, Represents a mapping vector, which is used to map the attention weight value. It is the hyperbolic tangent function;

[0021] To obtain the global power demand characteristics, the formula used is as follows:

[0022] ;

[0023] In the formula, represents the global power demand characteristics, Indicates the total number of time steps;

[0024] Design the power demand forecast activation function, the formula used is as follows:

[0025] ;

[0026] In the formula, represents the activation function for power demand prediction, represents the input variables for power demand forecast activation, It represents the nonlinear control parameter that can adjust the activation function. represents the weight distribution parameter;

[0027] The power demand forecast is generated using the following formula:

[0028] ;

[0029] In the formula, represents the power demand forecast result at the tth time step, represents the power demand forecast output weight matrix, represents the bias parameter of power demand forecast;

[0030] The power demand forecasting model training is specifically to train the power demand forecasting model using the historical power demand data in the preliminary data of charge and discharge control of the energy storage device to obtain a trained power demand forecasting model;

[0031] The regional real-time electricity demand forecast is specifically to use the real-time electricity demand data in the preliminary data of the energy storage device charging and discharging control as the input data of the trained electricity demand forecast model to obtain the regional real-time electricity demand forecast result, which includes the electricity demand, peak electricity demand value and valley electricity demand value.

[0032] Furthermore, the energy storage device charge and discharge prediction module is used to predict the future charge and discharge state of the energy storage device, including constructing an energy storage device charge and discharge prediction model, training the energy storage device charge and discharge prediction model, and predicting the real-time charge and discharge state of the energy storage device; specifically including the following steps:

[0033] Constructing a charging and discharging prediction model for energy storage equipment includes the following steps:

[0034] Extracting the long-term charging and discharging behavior characteristics of energy storage devices includes the following steps:

[0035] Perform a nonlinear transformation of the current layer using the following formula:

[0036] ;

[0037] In the formula, Represents the nonlinear transformation output of the current layer, Indicates that in the previous layer input sequence, the time step The input features of, i represents the current time step, k represents the index of the convolution kernel, Represents the expansion factor, dynamically adjusting the receptive field size of the convolution operation, represents the time weight, the weight parameter normalized by the softmax activation function, represents the weight parameter of the kth convolution kernel, represents the bias parameter of the kth convolution kernel, Represents the total number of convolution kernels;

[0038] Update the current layer output through the residual connection, the formula used is as follows:

[0039] ;

[0040] ;

[0041] In the formula, represents the dynamic weighting factor, represents the weight matrix of dynamic weighted training, represents the Sigmoid activation function, Represents the charging and discharging input characteristics of the energy storage device in the previous layer, Indicates the long-term charging and discharging behavior characteristics of the current layer of energy storage equipment, Indicates that The residual path input features after convolution processing, ReLU activation function.

[0042] The selection of key characteristics of energy storage device charging and discharging includes the following steps:

[0043] Calculate the global pooling value using the following formula:

[0044] ;

[0045] In the formula, represents the long-term charging and discharging behavior characteristics of the current layer energy storage device at the i-th time step, Represents the time step weight, which is used to emphasize the importance of key time steps. represents the total length of the time series, A global representation of the charge and discharge characteristics;

[0046] Expand the dynamic features, the formula used is as follows:

[0047] ;

[0048] In the formula, represents the output matrix of dynamic feature expansion, represents the linear feature weight matrix, represents the nonlinear feature weight matrix, Bias term parameter representing dynamic feature expansion;

[0049] The weighting mechanism of energy storage device charging and discharging characteristics is as follows:

[0050] ;

[0051] ;

[0052] ;

[0053] In the formula, Indicates the key characteristics of energy storage device charging and discharging. Represents the correction value, which is used to correct the attention score. Represents the result value of feature cross enhancement, represents the weight matrix used to generate the query matrix, represents the weight matrix used to generate the key matrix, represents the weight matrix used to generate the value matrix, Represents the fusion coefficient, which is used to adjust the weight ratio of the attention mechanism and cross enhancement. represents the dimension of the key matrix, represents the weight matrix of feature transformation, represents the bias term of feature transformation, represents the cross-feature matrix, represents the weight matrix of feature cross transformation, Represents the bias term of feature cross transformation;

[0054] Extract short-term dynamic charge and discharge characteristics, the formula used is as follows:

[0055] ;

[0056] ;

[0057] In the formula, represents the output of the update gate, Indicates the hidden state at the last moment. Represents the input data at the current moment, represents the weight matrix of the update gate, represents the bias parameter of the update gate, Represents the output of the reset gate, which controls the influence of past memory on the current moment. represents the weight matrix of the reset gate, Represents the bias parameter of the reset gate, Indicates the current state of the memory unit. represents the weight matrix of the current memory unit, Represents the bias parameter of the current memory unit, Indicates the hidden state at the current moment;

[0058] Dynamic feature fusion processing, the formula used is as follows:

[0059] ;

[0060] ;

[0061] In the formula, represents the feature fusion factor, represents the weight matrix of the joint mapping of output features, represents the bias parameter of the joint mapping of output features, represents the output feature after fusion at the tth time step, and Represents a linear transformation matrix, used to unify feature dimensions;

[0062] Design the energy storage device charge and discharge prediction activation function, the formula used is as follows:

[0063] ;

[0064] In the formula, represents the energy storage device charge and discharge prediction activation function, x represents the input variable of the energy storage device charge and discharge prediction activation function, A parameter representing the maximum magnitude of negative values, Indicates the sensitivity of adjusting positive value amplification;

[0065] Generate the prediction results of energy storage device charging and discharging, the formula used is as follows:

[0066] ;

[0067] In the formula, represents the prediction result of energy storage device charging and discharging at the tth time step, represents the output weight matrix, Represents the output bias parameter;

[0068] Energy storage device charge and discharge prediction model training, specifically using the historical energy storage device data in the energy storage device charge and discharge control preliminary data to train the energy storage device charge and discharge prediction model to obtain a trained energy storage device charge and discharge prediction model;

[0069] The real-time state prediction of energy storage device charging and discharging is specifically to use the real-time energy storage device data in the preliminary data of energy storage device charging and discharging control as the input data of the trained energy storage device charging and discharging prediction model to obtain the real-time state prediction result of energy storage device charging and discharging, and the real-time state prediction result of energy storage device charging and discharging is the charge state of energy storage device, the remaining energy of energy storage device and the ideal state energy storage energy.

[0070] Furthermore, the device charging and discharging strategy optimization module specifically obtains the optimal charging and discharging strategy of the energy storage device by improving the particle swarm optimization algorithm, including the following steps:

[0071] Initialization parameters, specifically, by constructing the initial parameters of the algorithm; the initial parameters of the algorithm include the number of particles N and the maximum number of iterations ;

[0072] The particle swarm is initialized, specifically by randomly generating particle positions and evaluating the fitness of the generated particles, wherein the particle positions represent a charging and discharging strategy of an energy storage device; specifically, the following steps are included:

[0073] Design particle position constraints, specifically including energy storage device charging and discharging power constraints and energy storage device remaining energy constraints; specifically including the following steps:

[0074] The charging and discharging power constraints of energy storage equipment use the following formula:

[0075] ;

[0076] In the formula, Represents the charging and discharging power of the energy storage device at time t, a positive value indicates discharging, and a negative value indicates charging; represents the peak power demand value, Indicates the low power demand value, and Indicates the grid regulation coefficient, ranging from , used to adjust the mandatory boundary of charge and discharge power. , the power grid needs energy storage equipment to discharge excessively to relieve peak pressure. , then the grid needs to limit the low-valley charging power to avoid over-consumption;

[0077] The remaining energy constraint of the energy storage device is used to ensure that the energy of the energy storage device is always not less than the remaining energy threshold in the future period; the formula used is as follows:

[0078] ;

[0079] In the formula, represents the remaining energy of the energy storage device at time t, represents the time window length, Indicates the total length of the time window, Indicates the remaining energy safety threshold of the energy storage device. represents the charging and discharging power of the energy storage device at time k, Represents the total capacity of the energy storage device, represents the power demand at time t, represents the power demand marketing coefficient, ranging from ;

[0080] Initialize the particle position. Specifically, the charging and discharging strategy of the energy storage device is the charging and discharging power of the energy storage device at each moment. The formula used is as follows:

[0081] ;

[0082] In the formula, represents the position of the i-th particle;

[0083] Particle fitness evaluation, specifically calculating the fitness value of particles in the particle swarm by designing a fitness function ; The design fitness function includes the energy storage device charging and discharging loss term, the grid demand matching term and the energy storage device remaining energy tracking term; the formula used is as follows:

[0084] ;

[0085] ;

[0086] In the formula, represents the fitness function, represents the loss parameter of energy storage equipment, represents the basic loss coefficient, represents the state of charge of the energy storage device at time t, Indicates the ideal storage energy of the energy storage device. Represents the balance weight parameter of the charging and discharging loss term of the energy storage device, represents the balancing weight parameter of the grid demand matching item, Represents the balance weight parameter of the remaining energy tracking item of the energy storage device;

[0087] Update the particle search parameters, specifically according to the nonlinear inertia weight strategy, the formula used is as follows:

[0088] ;

[0089] In the formula, Indicates The inertia weight of the iteration, represents the maximum value of the inertia weight, represents the minimum value of inertia weight, Indicates the current iteration number;

[0090] Update particle velocity and position using the following formula:

[0091] ;

[0092] In the formula, Indicates that the i-th particle is in the Iteration speed, Indicates that the i-th particle is in the Iteration speed, Indicates that the i-th particle is in the Iteration position, represents the local optimal position of individual particles, represents the global optimal position of the particle, and represents a random number in the range [0,1], represents the individual learning factor, which is used to control the speed at which particles move to the individual optimal position. represents the group learning factor, which is used to control the speed at which particles move to the global optimal position. represents the position of the i-th particle in the nt+1-th iteration;

[0093] Eliminate inferior particles, specifically if the fitness value of the current particle Less than the particle elimination threshold , then the particle is removed from the population and a particle that meets the constraints is randomly generated; the formula used is as follows:

[0094] ;

[0095] ;

[0096] In the formula, Represents the population fitness value of the current iteration, represents the average fitness of the population, Indicates the individual elimination coefficient of inferior particles;

[0097] Search determination, specifically, by constructing a search termination condition, searching and determining the global optimal position of the particle, and obtaining the data setting of the global optimal position of the particle;

[0098] The search termination conditions include threshold termination and iteration termination;

[0099] The threshold termination is specifically to set a fitness threshold, when the particle fitness value f i When it is above the fitness threshold, the search is completed;

[0100] The iteration termination specifically refers to terminating the iteration and obtaining the global optimal position of the particle when the maximum number of iterations is reached;

[0101] The global optimal position of the particle specifically refers to the optimal charging and discharging strategy of the energy storage device.

[0102] Furthermore, the energy storage device charging and discharging intelligent control module specifically inputs the real-time power demand forecast results of the region and the real-time status forecast results of the energy storage device charging and discharging into the equipment charging and discharging strategy optimization module, generates the optimal charging and discharging strategy for the energy storage device, and executes this strategy to realize intelligent charging and discharging control of the energy storage device.

[0103] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0104] (1) In response to the technical problems of low energy storage device life, low charging and discharging efficiency, and uncontrollable charging and discharging process in traditional energy storage device charging and discharging control systems, this solution innovatively proposes to dynamically adjust the charging and discharging strategy by real-time prediction of power demand and energy storage device status combined with optimization algorithms. It can flexibly select the best charging and discharging time for energy storage devices, ensure that energy storage devices are charged and discharged at the optimal time, maximize charging and discharging efficiency, reduce energy loss, effectively avoid overcharging and discharging, extend the service life of energy storage devices and reduce battery attenuation rate, improve the stability and responsiveness of the charging and discharging control system, solve the problems of power system overload and insufficient power of energy storage devices, and improve the operating efficiency of energy storage devices, thereby realizing intelligent control of energy storage devices in the region.

[0105] (2) In order to address the problems of insufficient feature extraction, inadequate modeling of time dependency, and limited ability to handle nonlinear features in existing models for predicting the charge and discharge of energy storage devices, which lead to inaccurate prediction of the charge and discharge status of energy storage devices, this solution innovatively proposes a method that combines long-term behavioral feature extraction with short-term dynamic charge and discharge feature modeling. By using a dynamic weighting mechanism and feature fusion strategy, this method solves the problem of insufficient capture of long-term trends and short-term fluctuation features by traditional models, and improves the model's adaptability and flexibility to complex energy storage device charge and discharge prediction scenarios, thereby improving the accuracy of the output results of energy storage device charge and discharge predictions and providing more data support for intelligent control of charge and discharge of energy storage devices.

[0106] (3) In view of the technical problem that the traditional equipment charging and discharging strategy optimization algorithm has a weak ability to obtain the global optimal solution, which leads to the final charging and discharging strategy being inappropriate, this scheme enhances the global optimal solution search and improves the flexibility of the algorithm by designing particle position constraints, fitness functions, nonlinear inertia weight strategies and strategies for eliminating inferior particles, so as to obtain the optimal charging and discharging strategy suitable for energy storage equipment and meet the intelligent charging and discharging control of energy storage equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0107] Figure 1 A schematic diagram of a module of an energy storage device charging and discharging control system based on artificial intelligence provided by the present invention;

[0108] Figure 2 It is a flow chart of the power demand forecasting module;

[0109] Figure 3 It is a flow chart of the charge and discharge prediction module of the energy storage device;

[0110] Figure 4 This is a flow chart of the device charging and discharging strategy optimization module;

[0111] Figure 5A schematic diagram of the process of constructing an energy storage device charge and discharge prediction model in an energy storage device charge and discharge prediction module;

[0112] Figure 6 This is a flow chart of particle swarm initialization in the device charging and discharging strategy optimization module;

[0113] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0114] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0115] In the description of the present invention, it is necessary to understand that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred system or element must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention.

[0116] Example 1, see Figure 1 , the present invention provides an energy storage device charging and discharging control system based on artificial intelligence, including a data acquisition module, a data preprocessing module, a power demand prediction module, an energy storage device charging and discharging prediction module, an equipment charging and discharging strategy optimization module and an energy storage device charging and discharging intelligent control module;

[0117] The data acquisition module acquires raw data of charge and discharge control of energy storage equipment by collecting data from the national power grid system, and sends the data to the data preprocessing module;

[0118] The data preprocessing module receives the data sent by the data acquisition module, performs data cleaning, time alignment, standardization processing and feature selection on the original data of the energy storage device charge and discharge control, obtains preliminary data of the energy storage device charge and discharge control, and sends the data to the power demand prediction module and the energy storage device charge and discharge prediction module;

[0119] The power demand prediction module receives the data sent by the data preprocessing module, builds a power demand prediction model, performs model training on the power demand prediction model, realizes real-time prediction of power demand, obtains regional real-time power demand prediction results, and sends the data to the energy storage device charging and discharging intelligent control module;

[0120] The energy storage device charge and discharge prediction module receives the data sent by the data preprocessing module, constructs an energy storage device charge and discharge prediction model, trains the energy storage device charge and discharge prediction model, realizes the real-time state prediction of the energy storage device charge and discharge, obtains the real-time state prediction result of the energy storage device charge and discharge, and sends the data to the energy storage device charge and discharge intelligent control module;

[0121] The device charge and discharge strategy optimization module receives data sent by the energy storage device charge and discharge intelligent control module, obtains the optimal charge and discharge strategy of the energy storage device by improving the particle swarm optimization algorithm, and sends the data to the energy storage device charge and discharge intelligent control module;

[0122] The energy storage device charge and discharge intelligent control module receives data sent by the power demand prediction module, the energy storage device charge and discharge prediction module and the device charge and discharge strategy optimization module, obtains the optimal charge and discharge strategy of the energy storage device by inputting the data into the device charge and discharge strategy optimization module, and executes this strategy to realize intelligent charge and discharge control of the energy storage device.

[0123] By executing the above operations, in response to the technical problems of low energy storage device life, low charging and discharging efficiency, and uncontrollable charging and discharging process in traditional energy storage device charging and discharging control systems, this solution innovatively proposes to predict power demand and energy storage device status in real time, and dynamically adjust the charging and discharging strategy in combination with optimization algorithms. It can flexibly select the best charging and discharging time for energy storage devices, ensure that energy storage devices are charged and discharged at the optimal time, maximize charging and discharging efficiency, reduce energy loss, effectively avoid over-charging and discharging, extend the service life of energy storage devices and reduce battery attenuation rate, improve the stability and responsiveness of the charging and discharging control system, solve the problems of power system overload and insufficient power of energy storage devices, and improve the operating efficiency of energy storage devices, thereby realizing intelligent control of energy storage devices in the region.

[0124] Example 2, see Figure 1, this embodiment is based on the above embodiment, the data acquisition module specifically acquires the original data of energy storage device charge and discharge control from the national power grid system by acquisition; the original data of energy storage device charge and discharge control includes historical power demand data, real-time power demand data, historical energy storage device data and real-time energy storage device data; the historical power demand data and real-time power demand data both include power load data, time characteristic data and weather data; the historical energy demand data also includes historical energy demand results; the power load data includes residential power load, industrial power load and commercial power load; the time characteristic data includes power demand rules in different time periods and changes in power demand during festivals; the weather data includes temperature, humidity, rainfall and special weather events; the historical energy storage device data and real-time energy storage device data both include device status data and device charge and discharge data; the device status data includes remaining power of energy storage device, degradation of energy storage device, service life of energy storage device, current data of energy storage device, voltage data of energy storage device and temperature of energy storage device; the device charge and discharge data includes the number of charge and discharge times of energy storage device, charge and discharge power of energy storage device and charge and discharge time of energy storage device.

[0125] Example 3, see Figure 1 This embodiment is based on the above embodiment, and the data preprocessing module specifically performs data cleaning, time alignment, standardization processing and feature selection on the original data of the energy storage device charge and discharge control to obtain preliminary data of the energy storage device charge and discharge control; the data cleaning is to process missing values, abnormal values ​​and repeated values; the time alignment is to unify the time granularity of the data through time grouping and aggregation, unify the time granularity to the hourly level, and align the timestamps of data from different sources; the standardization processing is based on the maximum and minimum normalization method to standardize the data; the feature selection is to use the correlation analysis method to screen out the relevant features with the power demand forecast and the energy storage device charge and discharge forecast.

[0126] Example 4, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. The power demand prediction module is used to predict the future power demand in the region, help the energy storage equipment to reasonably schedule the charging and discharging strategy, and avoid unnecessary high-cost operations, including power demand prediction model construction, power demand prediction model training and regional real-time power demand prediction; specifically includes the following steps:

[0127] The construction of the power demand forecasting model includes the following steps:

[0128] The formula used for extracting the time series features of power demand is as follows:

[0129] ;

[0130] In the formula, Indicates the unit runs the function, represents the hidden state at the kth time step, represents the input data at the kth time step, Indicates that it is The hidden state of time steps, Indicates that it is The cell state at time steps,

[0131] The attention weight is calculated using the following formula:

[0132] ;

[0133] In the formula, represents the attention weight value of the kth time step, represents the attention weight matrix, represents the attention bias parameter, Represents a mapping vector, which is used to map the attention weight value;

[0134] To obtain the global power demand characteristics, the formula used is as follows:

[0135] ;

[0136] In the formula, represents the global power demand characteristics, Indicates the total number of time steps;

[0137] Design the power demand forecast activation function, the formula used is as follows:

[0138] ;

[0139] In the formula, represents the activation function for power demand prediction, represents the input variables for power demand forecast activation, Represents the nonlinear control parameter that can adjust the activation function; represents the weight distribution parameter;

[0140] The power demand forecast is generated using the following formula:

[0141] ;

[0142] In the formula, represents the power demand forecast result at the tth time step, represents the power demand forecast output weight matrix, represents the bias parameter of power demand forecast;

[0143] The power demand forecasting model training is specifically to train the power demand forecasting model using the historical power demand data in the preliminary data of charge and discharge control of the energy storage device to obtain a trained power demand forecasting model;

[0144] The regional real-time electricity demand forecast is specifically to use the real-time electricity demand data in the preliminary data of the energy storage device charging and discharging control as the input data of the trained electricity demand forecast model to obtain the regional real-time electricity demand forecast result, which includes the electricity demand, peak electricity demand value and valley electricity demand value.

[0145] Example 5, see Figure 1 , Figure 3 and Figure 5 This embodiment is based on the above embodiment. The energy storage device charge and discharge prediction module is used to predict the future charge and discharge state of the energy storage device, including building an energy storage device charge and discharge prediction model, energy storage device charge and discharge prediction model training and energy storage device charge and discharge real-time state prediction; specifically includes the following steps:

[0146] Constructing a charging and discharging prediction model for energy storage equipment includes the following steps:

[0147] Extracting the long-term charging and discharging behavior characteristics of energy storage devices includes the following steps:

[0148] Perform a nonlinear transformation of the current layer using the following formula:

[0149] ;

[0150] In the formula, Represents the nonlinear transformation output of the current layer, Indicates that in the previous layer input sequence, the time step The input features of, i represents the current time step, k represents the index of the convolution kernel, Represents the expansion factor, dynamically adjusting the receptive field size of the convolution operation, represents the time weight, the weight parameter normalized by the softmax activation function, represents the weight parameter of the kth convolution kernel, represents the bias parameter of the kth convolution kernel, Represents the total number of convolution kernels;

[0151] Update the current layer output through the residual connection, the formula used is as follows:

[0152] ;

[0153] ;

[0154] In the formula, represents the dynamic weighting factor, represents the weight matrix of dynamic weighted training, represents the Sigmoid function, Represents the charging and discharging input characteristics of the energy storage device in the previous layer, Indicates the long-term trend characteristics of the current layer of energy storage equipment charging and discharging, Indicates that The residual path input features after convolution processing, ReLU activation function.

[0155] The selection of key characteristics of energy storage device charging and discharging includes the following steps:

[0156] Calculate the global pooling value using the following formula:

[0157] ;

[0158] In the formula, represents the long-term charging and discharging behavior characteristics of the current layer energy storage device at the i-th time step, Represents the time step weight, which is used to emphasize the importance of key time steps. represents the total length of the time series, A global representation of the charge and discharge characteristics;

[0159] Expand the dynamic features, the formula used is as follows:

[0160] ;

[0161] In the formula, represents the output matrix of dynamic feature expansion, represents the linear feature weight matrix, represents the nonlinear feature weight matrix, Bias term parameter representing dynamic feature expansion;

[0162] The weighting mechanism of energy storage device charging and discharging characteristics is as follows:

[0163] ;

[0164] ;

[0165] ;

[0166] In the formula, Indicates the key characteristics of energy storage device charging and discharging. Represents the correction value, which is used to correct the attention score. Represents the result value of feature cross enhancement, represents the weight matrix used to generate the query matrix, represents the weight matrix used to generate the key matrix, represents the weight matrix used to generate the value matrix, Represents the fusion coefficient, which is used to adjust the weight ratio of the attention mechanism and cross enhancement. represents the dimension of the key matrix, represents the weight matrix of feature transformation, represents the bias term of feature transformation, represents the cross-feature matrix, which is the cross-feature obtained by combining the temperature, current and voltage features of the energy storage device. represents the weight matrix of feature cross transformation, Represents the bias term of feature cross transformation;

[0167] Extract short-term dynamic charge and discharge characteristics, the formula used is as follows:

[0168] ;

[0169] ;

[0170] In the formula, represents the output of the update gate, Indicates the hidden state at the last moment. Represents the input data at the current moment, represents the weight matrix of the update gate, represents the bias parameter of the update gate, Represents the output of the reset gate, which controls the influence of past memory on the current moment. represents the weight matrix of the reset gate, Represents the bias parameter of the reset gate, Indicates the current state of the memory unit. represents the weight matrix of the current memory unit, Represents the bias parameter of the current memory unit, represents the hyperbolic tangent function, Indicates the hidden state at the current moment;

[0171] Dynamic feature fusion processing, the formula used is as follows:

[0172] ;

[0173] ;

[0174] In the formula, represents the feature fusion factor, represents the weight matrix of the joint mapping of output features, represents the bias parameter of the joint mapping of output features, represents the output feature after fusion at the tth time step, and Represents a linear transformation matrix, used to unify feature dimensions;

[0175] Design the energy storage device charge and discharge prediction activation function, the formula used is as follows:

[0176] ;

[0177] In the formula, represents the energy storage device charge and discharge prediction activation function, x represents the input variable of the energy storage device charge and discharge prediction activation function, A parameter representing the maximum magnitude of negative values, Indicates the sensitivity of adjusting positive value amplification;

[0178] Generate the prediction results of energy storage device charging and discharging, the formula used is as follows:

[0179] ;

[0180] In the formula, represents the prediction result of energy storage device charging and discharging at the tth time step, represents the output weight matrix, Represents the output bias parameter;

[0181] Energy storage device charge and discharge prediction model training, specifically using the historical energy storage device data in the energy storage device charge and discharge control preliminary data to train the energy storage device charge and discharge prediction model to obtain a trained energy storage device charge and discharge prediction model;

[0182] The real-time state prediction of energy storage device charging and discharging is specifically to use the real-time energy storage device data in the preliminary data of energy storage device charging and discharging control as the input data of the trained energy storage device charging and discharging prediction model to obtain the real-time state prediction result of energy storage device charging and discharging, and the real-time state prediction result of energy storage device charging and discharging is the charge state of energy storage device, the remaining energy of energy storage device and the ideal state energy storage energy.

[0183] By performing the above operations, this solution innovatively proposes a method that combines long-term behavior feature extraction with short-term dynamic charge and discharge feature modeling to address the problems of insufficient feature extraction, insufficient time dependency modeling, and limited ability to handle nonlinear features in existing models for predicting the charge and discharge of energy storage devices, which leads to inaccurate prediction of the charge and discharge status of energy storage devices. It uses a dynamic weighting mechanism and feature fusion strategy to solve the problem of insufficient capture of long-term trends and short-term fluctuation features by traditional models, and improves the model's adaptability and flexibility to complex energy storage device charge and discharge prediction scenarios, thereby improving the accuracy of the output results of energy storage device charge and discharge predictions and providing more data support for intelligent control of charge and discharge of energy storage devices.

[0184] Example 6, see Figure 1 , Figure 4 and Figure 6 This embodiment is based on the above embodiment. The device charging and discharging strategy optimization module specifically obtains the optimal charging and discharging strategy of the energy storage device by improving the particle swarm optimization algorithm, including the following steps:

[0185] Initialization parameters, specifically, by constructing the initial parameters of the algorithm; the initial parameters of the algorithm include the number of particles N and the maximum number of iterations ;

[0186] The particle swarm is initialized, specifically by randomly generating particle positions and evaluating the fitness of the generated particles, wherein the particle positions represent a charging and discharging strategy of an energy storage device; specifically, the following steps are included:

[0187] The particle position constraints are designed to ensure the rationality and feasibility of the charging and discharging strategy and avoid invalid and unreasonable charging and discharging schemes, including the charging and discharging power constraints of the energy storage device and the remaining energy constraints of the energy storage device; the specific steps include:

[0188] The charging and discharging power constraints of energy storage equipment use the following formula:

[0189] ;

[0190] In the formula, Represents the charging and discharging power of the energy storage device at time t, a positive value indicates discharging, and a negative value indicates charging; represents the peak power demand value, Indicates the low power demand value, and Indicates the grid regulation coefficient, ranging from , used to adjust the mandatory boundary of charge and discharge power. , the power grid needs energy storage equipment to discharge excessively to relieve peak pressure. , then the grid needs to limit the low-valley charging power to avoid over-consumption;

[0191] The remaining energy constraint of the energy storage device is used to ensure that the energy of the energy storage device is always not less than the remaining energy threshold in the future period; the formula used is as follows:

[0192] ;

[0193] In the formula, represents the remaining energy of the energy storage device at time t, represents the time window length, Indicates the total length of the time window, Indicates the remaining energy safety threshold of the energy storage device. represents the charging and discharging power of the energy storage device at time k, Represents the total capacity of the energy storage device, represents the power demand at time t, represents the power demand marketing coefficient, ranging from ;

[0194] Initialize the particle position. Specifically, the charging and discharging strategy of the energy storage device is the charging and discharging power of the energy storage device at each moment. The formula used is as follows:

[0195] ;

[0196] In the formula, represents the position of the i-th particle;

[0197] Particle fitness evaluation is used to enable particles to more accurately measure the pros and cons of strategies, thereby more effectively guiding the search process and obtaining the optimal charging and discharging strategy for energy storage devices. Specifically, the fitness value of particles in the particle swarm is calculated by designing a fitness function. ; The design fitness function includes the energy storage device charging and discharging loss term, the grid demand matching term and the energy storage device remaining energy tracking term; the formula used is as follows:

[0198] ;

[0199] ;

[0200] In the formula, represents the fitness function, represents the loss parameter of energy storage equipment, represents the basic loss coefficient, represents the state of charge of the energy storage device at time t, Indicates the ideal storage energy of the energy storage device. Represents the balance weight parameter of the charging and discharging loss term of the energy storage device, represents the balancing weight parameter of the grid demand matching item, Represents the balance weight parameter of the remaining energy tracking item of the energy storage device;

[0201] Update the particle search parameters, specifically according to the nonlinear inertia weight strategy, the formula used is as follows:

[0202] ;

[0203] In the formula, Indicates The inertia weight of the iteration, represents the maximum value of the inertia weight, represents the minimum value of inertia weight, Indicates the current iteration number;

[0204] Update particle velocity and position using the following formula:

[0205] ;

[0206] In the formula, Indicates that the i-th particle is in the Iteration speed, Indicates that the i-th particle is in the Iteration speed, Indicates that the i-th particle is in the Iteration position, represents the local optimal position of individual particles, represents the global optimal position of the particle, and represents a random number in the range [0,1], represents the individual learning factor, which is used to control the speed at which particles move to the individual optimal position. represents the group learning factor, which is used to control the speed at which particles move to the global optimal position. represents the position of the i-th particle in the nt+1-th iteration;

[0207] Eliminate inferior particles, which are used to eliminate particles with poor performance, improve the overall quality of the population, speed up the convergence speed, and obtain the optimal charging and discharging strategy of the energy storage device more quickly. Specifically, if the fitness value of the current particle Less than the particle elimination threshold , then the particle is removed from the population and a particle that meets the constraints is randomly generated; the formula used is as follows:

[0208] ;

[0209] ;

[0210] In the formula, Represents the population fitness value of the current iteration, represents the average fitness of the population, Indicates the individual elimination coefficient of inferior particles;

[0211] Search determination, specifically, by constructing a search termination condition, searching and determining the global optimal position of the particle, and obtaining the data setting of the global optimal position of the particle;

[0212] The search termination conditions include threshold termination and iteration termination;

[0213] The threshold termination is specifically to set a fitness threshold, when the particle fitness value f i When it is above the fitness threshold, the search is completed;

[0214] The iteration termination specifically refers to terminating the iteration and obtaining the global optimal position of the particle when the maximum number of iterations is reached;

[0215] The global optimal position of the particle specifically refers to the optimal charging and discharging strategy of the energy storage device.

[0216] By performing the above operations, the traditional equipment charging and discharging strategy optimization algorithm has a weak ability to obtain the global optimal solution, which leads to the technical problem that the final charging and discharging strategy is not suitable. This solution enhances the global optimal solution search and improves the flexibility of the algorithm by designing particle position constraints, fitness functions, nonlinear inertia weight strategies and eliminating inferior particles, so as to obtain the optimal charging and discharging strategy suitable for energy storage equipment and meet the intelligent charging and discharging control of energy storage equipment.

[0217] Embodiment 7, see Figure 1 This embodiment is based on the above embodiment. The energy storage device charging and discharging intelligent control module specifically inputs the real-time power demand forecast results of the region and the real-time state forecast results of the energy storage device charging and discharging into the device charging and discharging strategy optimization module, generates the optimal charging and discharging strategy for the energy storage device, and executes this strategy to realize the intelligent charging and discharging control of the energy storage device.

[0218] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0219] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0220] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. An energy storage device charging and discharging control system based on artificial intelligence, characterized in that: It includes data acquisition module, data preprocessing module, power demand prediction module, energy storage equipment charging and discharging prediction module, equipment charging and discharging strategy optimization module and energy storage equipment charging and discharging intelligent control module; The data acquisition module acquires the original data of charge and discharge control of energy storage equipment by collecting data from the national power grid system; The data preprocessing module performs data cleaning, time alignment, standardization processing and feature selection on the original data of the energy storage device charge and discharge control to obtain preliminary data of the energy storage device charge and discharge control; The power demand forecasting module is used to forecast the future power demand in the region, obtain the global power demand characteristics through power demand time series feature extraction and attention weight calculation to build a power demand forecasting model, train the power demand forecasting model, realize real-time power demand forecasting, and obtain the real-time power demand forecasting result of the region; The energy storage device charge and discharge prediction module constructs an energy storage device charge and discharge prediction model, trains the energy storage device charge and discharge prediction model, realizes the real-time state prediction of the energy storage device charge and discharge, and obtains the real-time state prediction result of the energy storage device charge and discharge; the construction of the energy storage device charge and discharge prediction model includes the following steps: extracting the long-term behavior characteristics of the energy storage device charge and discharge, selecting the key characteristics of the energy storage device charge and discharge, extracting the short-term dynamic charge and discharge characteristics, fusion processing of dynamic characteristics, designing the energy storage device charge and discharge prediction activation function, and generating the energy storage device charge and discharge prediction result; The device charging and discharging strategy optimization module improves the particle swarm optimization algorithm for obtaining the optimal charging and discharging strategy of the energy storage device by designing particle position constraints, designing fitness functions, nonlinear inertia weight strategies, and eliminating inferior particle strategies, thereby obtaining the optimal charging and discharging strategy of the energy storage device; The energy storage device charge and discharge intelligent control module inputs data into the device charge and discharge strategy optimization module to obtain the optimal charge and discharge strategy of the energy storage device, and executes this strategy to realize intelligent charge and discharge control of the energy storage device.

2. According to claim 1, an energy storage device charging and discharging control system based on artificial intelligence is characterized in that: The power demand prediction module is used to predict future power demand in the region, help energy storage equipment to reasonably schedule charging and discharging strategies, and avoid unnecessary high-cost operations, including power demand prediction model construction, power demand prediction model training, and regional real-time power demand prediction; specifically, the following steps are included: The construction of the power demand forecasting model includes the following steps: The formula used for extracting the time series features of power demand is as follows: ； In the formula, Indicates the unit runs the function, represents the hidden state at the kth time step, represents the input data at the kth time step, Indicates that it is The hidden state of time steps, Indicates that it is The cell state at time steps, The attention weight is calculated using the following formula: ； In the formula, represents the attention weight value of the kth time step, represents the attention weight matrix, represents the attention bias parameter, Represents a mapping vector, which is used to map the attention weight value. It is the hyperbolic tangent function; To obtain the global power demand characteristics, the formula used is as follows: ； In the formula, represents the global power demand characteristics, Indicates the total number of time steps; Design the power demand forecast activation function, the formula used is as follows: ； In the formula, represents the activation function for power demand prediction, represents the input variables for power demand forecast activation, It indicates that the nonlinear control parameters of the activation function can be adjusted. represents the weight distribution parameter; The power demand forecast is generated using the following formula: ； In the formula, represents the power demand forecast result at the tth time step, represents the power demand forecast output weight matrix, represents the bias parameter of power demand forecast; The power demand forecasting model training is specifically to train the power demand forecasting model using the historical power demand data in the preliminary data of charge and discharge control of the energy storage device to obtain a trained power demand forecasting model; The regional real-time electricity demand forecast is specifically to use the real-time electricity demand data in the preliminary data of the energy storage device charging and discharging control as the input data of the trained electricity demand forecast model to obtain the regional real-time electricity demand forecast result, which includes the electricity demand, peak electricity demand value and valley electricity demand value.

3. According to claim 1, an energy storage device charge and discharge control system based on artificial intelligence is characterized in that: The energy storage device charge and discharge prediction module is used to predict the future charge and discharge state of the energy storage device, including building an energy storage device charge and discharge prediction model, energy storage device charge and discharge prediction model training and energy storage device charge and discharge real-time state prediction; specifically including the following steps: Constructing a charging and discharging prediction model for energy storage equipment includes the following steps: Extracting the long-term charging and discharging behavior characteristics of energy storage devices includes the following steps: Perform a nonlinear transformation of the current layer using the following formula: ； In the formula, Represents the nonlinear transformation output of the current layer, Indicates that in the previous layer input sequence, the time step The input features of, i represents the current time step, k represents the index of the convolution kernel, Represents the expansion factor, dynamically adjusting the receptive field size of the convolution operation, represents the time weight, the weight parameter normalized by the softmax activation function, represents the weight parameter of the kth convolution kernel, represents the bias parameter of the kth convolution kernel, Represents the total number of convolution kernels; Update the current layer output through the residual connection, the formula used is as follows: ； ； In the formula, represents the dynamic weighting factor, represents the weight matrix of dynamic weighted training, represents the Sigmoid activation function, Represents the charging and discharging input characteristics of the energy storage device in the previous layer, Indicates the long-term charging and discharging behavior characteristics of the current layer of energy storage equipment, Indicates that The residual path input features after convolution processing, ReLU activation function. The selection of key characteristics of energy storage device charging and discharging includes the following steps: Calculate the global pooling value using the following formula: ； In the formula, represents the long-term charging and discharging behavior characteristics of the current layer of energy storage equipment at the i-th time step, Represents the time step weight, which is used to emphasize the importance of key time steps. represents the total length of the time series, A global representation of the charge and discharge characteristics; Expand the dynamic features, the formula used is as follows: ； In the formula, represents the output matrix of dynamic feature expansion, represents the linear feature weight matrix, represents the nonlinear feature weight matrix, Bias term parameter representing dynamic feature expansion; The weighting mechanism of energy storage device charging and discharging characteristics is as follows: ； ； ； In the formula, Indicates the key characteristics of energy storage device charging and discharging. Represents the correction value, which is used to correct the attention score. Represents the result value of feature cross enhancement, represents the weight matrix used to generate the query matrix, represents the weight matrix used to generate the key matrix, represents the weight matrix used to generate the value matrix, Represents the fusion coefficient, which is used to adjust the weight ratio of the attention mechanism and cross enhancement. represents the dimension of the key matrix, represents the weight matrix of feature transformation, represents the bias term of feature transformation, represents the cross-feature matrix, represents the weight matrix of feature cross transformation, Represents the bias term of feature cross transformation; Extract short-term dynamic charge and discharge characteristics, the formula used is as follows: ； ； In the formula, represents the output of the update gate, Indicates the hidden state at the last moment. Represents the input data at the current moment, represents the weight matrix of the update gate, represents the bias parameter of the update gate, Represents the output of the reset gate, which controls the influence of past memory on the current moment. represents the weight matrix of the reset gate, Represents the bias parameter of the reset gate, Indicates the current state of the memory unit. represents the weight matrix of the current memory unit, Represents the bias parameter of the current memory unit, Indicates the hidden state at the current moment; Dynamic feature fusion processing, the formula used is as follows: ； ； In the formula, represents the feature fusion factor, represents the weight matrix of the joint mapping of output features, represents the bias parameter of the joint mapping of output features, represents the output feature after fusion at the tth time step, and Represents a linear transformation matrix, used to unify feature dimensions; Design the energy storage device charge and discharge prediction activation function, the formula used is as follows: ； In the formula, represents the energy storage device charge and discharge prediction activation function, x represents the input variable of the energy storage device charge and discharge prediction activation function, A parameter representing the maximum magnitude of negative values, Indicates the sensitivity of adjusting positive value amplification; Generate the prediction results of energy storage device charging and discharging, the formula used is as follows: ； In the formula, represents the prediction result of energy storage device charging and discharging at the tth time step, represents the output weight matrix, Represents the output bias parameter; Energy storage device charge and discharge prediction model training, specifically using the historical energy storage device data in the energy storage device charge and discharge control preliminary data to train the energy storage device charge and discharge prediction model to obtain a trained energy storage device charge and discharge prediction model; The real-time state prediction of energy storage device charging and discharging is specifically to use the real-time energy storage device data in the preliminary data of energy storage device charging and discharging control as the input data of the trained energy storage device charging and discharging prediction model to obtain the real-time state prediction result of energy storage device charging and discharging, and the real-time state prediction result of energy storage device charging and discharging is the charge state of energy storage device, the remaining energy of energy storage device and the ideal state energy storage energy.

4. The artificial intelligence-based energy storage device charge and discharge control system according to claim 1, characterized in that: The device charging and discharging strategy optimization module specifically obtains the optimal charging and discharging strategy of the energy storage device by improving the particle swarm optimization algorithm, including the following steps: Initialization parameters, specifically, by constructing the initial parameters of the algorithm; the initial parameters of the algorithm include the number of particles N and the maximum number of iterations ; The particle swarm is initialized, specifically by randomly generating particle positions and evaluating the fitness of the generated particles, wherein the particle positions represent a charging and discharging strategy of an energy storage device; specifically, the following steps are included: Design particle position constraints, specifically including energy storage device charging and discharging power constraints and energy storage device remaining energy constraints; specifically including the following steps: The charging and discharging power constraints of energy storage equipment use the following formula: ； In the formula, Represents the charging and discharging power of the energy storage device at time t, a positive value indicates discharging, and a negative value indicates charging; represents the peak power demand value, Indicates the valley power demand value, and Indicates the grid regulation coefficient, ranging from , used to adjust the mandatory boundary of charge and discharge power. , the power grid needs energy storage equipment to discharge excessively to relieve peak pressure. , then the grid needs to limit the low-valley charging power to avoid over-consumption; The remaining energy constraint of the energy storage device is used to ensure that the energy of the energy storage device is always not less than the remaining energy threshold in the future period; the formula used is as follows: ； In the formula, represents the remaining energy of the energy storage device at time t, represents the time window length, Indicates the total length of the time window, Indicates the remaining energy safety threshold of the energy storage device. represents the charging and discharging power of the energy storage device at time k, Represents the total capacity of the energy storage device, represents the power demand at time t, represents the power demand marketing coefficient, ranging from ; Initialize the particle position. Specifically, the charging and discharging strategy of the energy storage device is the charging and discharging power of the energy storage device at each moment. The formula used is as follows: ； In the formula, represents the position of the i-th particle; Particle fitness evaluation, specifically calculating the fitness value of particles in the particle swarm by designing a fitness function ; The design fitness function includes the energy storage device charging and discharging loss term, the grid demand matching term and the energy storage device remaining energy tracking term; the formula used is as follows: ； ； In the formula, represents the fitness function, represents the loss parameter of energy storage equipment, represents the basic loss coefficient, represents the state of charge of the energy storage device at time t, Indicates the ideal storage energy of the energy storage device. Represents the balance weight parameter of the charging and discharging loss term of the energy storage device, represents the balancing weight parameter of the grid demand matching item, Represents the balance weight parameter of the remaining energy tracking item of the energy storage device; Update the particle search parameters, specifically according to the nonlinear inertia weight strategy, the formula used is as follows: ； In the formula, Indicates The inertia weight of the iteration, represents the maximum value of the inertia weight, represents the minimum value of inertia weight, Indicates the current iteration number; Update particle velocity and position using the following formula: ； In the formula, Indicates that the i-th particle is in the Iteration speed, Indicates that the i-th particle is in Iteration speed, Indicates that the i-th particle is in Iteration position, represents the local optimal position of individual particles, represents the global optimal position of the particle, and represents a random number in the range [0,1], represents the individual learning factor, which is used to control the speed at which particles move to the individual optimal position. represents the group learning factor, which is used to control the speed at which particles move to the global optimal position. represents the position of the i-th particle in the nt+1-th iteration; Eliminate inferior particles, specifically if the fitness value of the current particle Less than the particle elimination threshold , then the particle is removed from the population and a particle that meets the constraints is randomly generated; the formula used is as follows: ； ； In the formula, Represents the population fitness value of the current iteration, represents the average fitness of the population, Indicates the individual elimination coefficient of inferior particles; Search determination, specifically, by constructing a search termination condition, searching and determining the global optimal position of the particle, and obtaining the data setting of the global optimal position of the particle; The search termination conditions include threshold termination and iteration termination; The threshold termination is specifically to set a fitness threshold, when the particle fitness value f i When it is above the fitness threshold, the search is completed; The iteration termination specifically refers to terminating the iteration and obtaining the global optimal position of the particle when the maximum number of iterations is reached; The global optimal position of the particle specifically refers to the optimal charging and discharging strategy of the energy storage device.

5. The energy storage device charge and discharge control system based on artificial intelligence according to claim 1, characterized in that: The energy storage device charging and discharging intelligent control module specifically inputs the real-time power demand forecast results of the region and the real-time state forecast results of the energy storage device charging and discharging into the device charging and discharging strategy optimization module, generates the optimal charging and discharging strategy for the energy storage device, and executes this strategy to realize intelligent charging and discharging control of the energy storage device.

6. The energy storage device charge and discharge control system based on artificial intelligence according to claim 1, characterized in that: The data acquisition module specifically acquires the original data of energy storage device charge and discharge control from the national power grid system by acquisition; the original data of energy storage device charge and discharge control includes historical power demand data, real-time power demand data, historical energy storage device data and real-time energy storage device data; the historical power demand data and real-time power demand data both include power load data, time characteristic data and weather data; the historical energy demand data also includes historical energy demand results; the historical energy storage device data and real-time energy storage device data both include device status data and device charge and discharge data.

7. The artificial intelligence-based energy storage device charge and discharge control system according to claim 1, characterized in that: The data preprocessing module specifically performs data cleaning, time alignment, standardization and feature selection on the original data of the energy storage device charge and discharge control to obtain preliminary data of the energy storage device charge and discharge control; the data cleaning is to process missing values, abnormal values ​​and repeated values; the time alignment is to unify the time granularity of the data through time grouping and aggregation, unify the time granularity to the hourly level, and align the timestamps of data from different sources; the standardization is to standardize the data based on the maximum and minimum normalization method; the feature selection is to use the correlation analysis method to screen out the relevant features with the power demand forecast and the energy storage device charge and discharge forecast.

Images