Peak clipping and valley filling type energy storage battery station deployment management and control system

Abstract

The invention discloses a peak clipping and valley filling type energy storage battery station allocation management and control system, and relates to the technical field of energy storage battery base station management, and the key points of the technical scheme are that the system comprises a cloud end scheduling center, an edge control end and an execution terminal; the execution terminal comprises a monitoring module, an energy storage battery monitoring module, a data sending module and an energy storage battery control module. According to the energy storage battery station regulation and control system, on the basis of original data judgment, the change of the concentration of vanadium ions with different valence states in the redox process of the vanadium ions in battery liquid in an energy storage battery is monitored when the energy storage battery is charged and discharged, and the concentration of the vanadium ions in the battery liquid in the energy storage battery is judged according to the concentration change condition. The state (charging and discharging) of the energy storage battery at the moment is directly and truly judged, meanwhile, the electric power of the energy storage battery can be accurately judged in combination with corrected charging and discharging data of the energy storage battery, and the loss, efficiency and aging degree of the battery can be truly and accurately judged.

Description

Technical Field

[0001] This invention relates to the field of energy storage battery base station management technology, and more specifically, to a peak-shaving and valley-filling energy storage battery station dispatching and control system. Background Technology

[0002] Peak shaving and valley filling refer to the process by which power companies, through necessary technical and management means, combined with some administrative measures, reduce peak loads and increase off-peak loads, smooth the load curve, increase load factor, reduce electricity demand, reduce generator investment, and stabilize grid operation.

[0003] In power systems, "peak shaving and valley filling" is an important means of balancing the peak and valley differences of the power grid and improving operating efficiency by adjusting the electricity load and energy storage system. During peak electricity demand, the energy storage system balances the electricity load by discharging the energy storage battery. During off-peak electricity demand, the energy storage system supplements the power by storing electricity. Due to its unique chemical properties, engineering advantages and compatibility with large-scale energy storage needs, the batteries used in energy storage battery stations are generally vanadium redox flow batteries.

[0004] Current energy storage systems are mainly based on peak shaving and valley filling periods. They monitor and determine relevant data such as electricity load, energy storage cluster power status (discharging and charging), and power loss cost data to control and manage energy storage batteries during peak shaving and valley filling periods. However, during charging and discharging, the power transmission of energy storage batteries is affected by factors such as battery aging, battery fluid deterioration, battery positive and negative terminal hardware (aging and connection), and the environment during power transmission. All of these factors cause power loss, resulting in errors between the data on the power flow of the energy storage batteries and the power generated by the batteries themselves. This leads to inaccuracies in the judgment of the dispatch management system and the power allocation of the energy storage battery stations.

[0005] Therefore, in order to solve the above-mentioned technical problems, this application proposes a peak-shaving and valley-filling energy storage battery station dispatch and control system. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the present invention aims to provide a peak-shaving and valley-filling energy storage battery station dispatching and control system.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a peak-shaving and valley-filling energy storage battery station dispatching and management system, comprising: a cloud dispatching center, an edge control terminal, and an execution terminal;

[0008] The execution terminal includes a monitoring module, an energy storage battery monitoring module, a data transmission module, and an energy storage battery control module;

[0009] The monitoring module is used to monitor power fluctuations in the distributed energy storage cluster and determine peak shaving times.

[0010] The system includes a peak shaving reference point and a valley filling reference point; the energy storage battery monitoring module monitors the dynamic changes in the concentration of vanadium ions at different valence states at the positive and negative electrodes in the battery solution during charging and discharging, as well as the operating rate of the liquid pump under the dynamic changes in vanadium ions; the data transmission module transmits power fluctuation data, dynamic change image data of vanadium ion concentration at different valence states, and the operating rate of the liquid pump under the dynamic changes in vanadium ions at the peak shaving reference point and the valley filling reference point; and the energy storage battery control module receives instructions to control the charging and discharging of the energy storage battery.

[0011] The cloud-based dispatch center includes a data receiving module, a data preprocessing module, and a computing module;

[0012] The data receiving module receives power fluctuation data at peak shaving and valley filling reference points, dynamic change image data of vanadium ion concentrations at different valence states, power grid data, and the operating rate of the liquid pump under dynamic changes in vanadium ions from the execution terminal. The data preprocessing module corrects the power fluctuation data at peak shaving and valley filling reference points, the dynamic change image data of vanadium ion concentrations at different valence states, and the operating rate of the liquid pump. The calculation module constructs a comprehensive model based on the corrected power fluctuation data and image data, combined with the power grid transmission data, the dynamic change image data of vanadium ion concentrations at different valence states, and the data corrected by the liquid pump operating rate, and formulates a power allocation plan based on the model.

[0013] The edge control terminal includes a receiving control module, which receives instructions from the cloud dispatch center, distributes power to each energy storage station within the adjustment area, and sends execution instructions to the execution terminal during peak shaving and valley filling periods.

[0014] Furthermore, the dynamic changes in the concentration of vanadium ions at different valence states at the positive and negative electrodes in the battery solution were investigated using ultraviolet-visible spectrophotometry.

[0015] Furthermore, the power fluctuation data and liquid pump operating rate data at the peak shaving period reference point and the valley filling period reference point are combined with the time series data and the LSTM model St= Make corrections;

[0016] Where St is the hidden state at the current time step t; The input vector at the current time step t; The weight matrix is ​​input into the hidden state; This is the hidden state of the previous time step t-1; is the weight matrix from hidden state to hidden state; f is the activation function.

[0017] Furthermore, the dynamic changes in vanadium ion concentrations at different valence states are corrected using image data through a combination of CNN loss function and model training. The loss function is as follows:

[0018] ;

[0019] in, The loss function; For this is the first The true label of each sample; For this is the first The predicted probability of a sample.

[0020] Furthermore, the integrated model is generated based on power fluctuation data from power grid transmission data, peak shaving period reference points and valley filling period reference points, dynamic change image data model of vanadium ion concentration in different valence states, and liquid pump operating rate model.

[0021] The models for power fluctuation data at peak-shaving and valley-filling time reference points use the peak-shaving time reference point as the starting point and the valley-filling time reference point as the ending point. The models are constructed by combining the changes in charging and discharging power of the energy storage battery with battery loss costs throughout the entire time period. The objective function is as follows:

[0022] Where Pbat is the energy storage charging and discharging power (positive for discharging, negative for charging); ΔPref is the deviation between the reference point and the current load; and Dbat is the battery loss cost.

[0023] Furthermore, the dynamic change image data model of vanadium ion concentration in different valence states is based on the conservation of momentum of vanadium ions in different valence states in the energy storage battery fluid during charging and discharging. and The concentration change;

[0024] That is, when charging, Oxidized In the positive electrode electrolyte Concentration increases Concentration decreases; during discharge, Restored to , Concentration decreases, As concentration increases, dynamic image features are extracted, and the Lucas-Kanades algorithm combined with a mathematical model is used to calculate the formula: Build.

[0025] Furthermore, the operating rate model of the liquid pump is based on flow characteristics and pump dynamics, and is constructed using a first-order linear differential equation. The specific formula is as follows:

[0026] ;

[0027] in, This is the pump's gain coefficient; This refers to the pump's input pressure. This refers to the pump's output pressure.

[0028] Furthermore, the constraints for the model of power fluctuation data at the peak-shaving and valley-filling time reference points include the following: ,in In order to be in The state of the battery at time (charging, discharging); In time The state of the battery at that time (charging, discharging);

[0029] For charge and discharge efficiency; This refers to the battery's rated capacity. In time Battery power at that time; For time step.

[0030] A peak-shaving and valley-filling energy storage battery station dispatching and management method specifically includes the following:

[0031] Collect power grid data and charging and discharging data of each energy storage battery in the distributed energy storage cluster to determine the benchmark points for peak shaving and valley filling periods. Collect fluctuation data of each energy storage battery based on power load within the benchmark points for peak shaving and valley filling periods. After correction, upload the data to the cloud dispatch center. Combine the battery loss cost data to build a model. Through model evolution, generate model A.

[0032] During peak shaving and valley filling periods, data were collected at reference points for each energy storage battery during charging and discharging, including the concentrations of electrolyte in the positive and negative electrodes. and The dynamically changing image data and the data on the operating rate of the liquid pump during the charging and discharging of the energy storage battery are corrected and uploaded to the cloud dispatch center to build a model. Through model evolution, model B is generated.

[0033] Model A and Model B are evolved to generate Model C. Based on the power load during peak shaving and valley filling periods, Model C can predict the charging and discharging of each energy storage battery in the energy storage cluster and send power allocation instructions to the edge control terminal. During peak shaving and valley filling periods, execution instructions are sent to the execution terminal to precisely control the charging and discharging of each energy storage battery in the energy storage cluster.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1. This energy storage battery station control system, based on the original data judgment, adds monitoring of the changes in vanadium ion concentration in the battery fluid during the redox process of the energy storage battery. By observing the changes in vanadium ion concentration in different valence states at the positive and negative terminals of the energy storage battery, the system can directly and accurately determine the current state (charging and discharging) of the energy storage battery. At the same time, it can also combine the corrected charging and discharging data of the energy storage battery to accurately determine the power of the energy storage battery, thereby enabling the control system to accurately allocate the power of the energy storage battery station during peak shaving and valley filling periods.

[0036] 2. This energy storage battery station control system monitors the changes in vanadium ion concentration in the battery fluid within the energy storage battery. This not only enables precise power distribution within the energy storage battery station but also allows for accurate and reliable assessment of the battery's own wear, efficiency, and aging, achieving real-time monitoring of the energy storage battery's condition. Attached Figure Description

[0037] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0038] Figure 1 This is a flowchart of the energy storage battery station dispatch and control system in this invention. Detailed Implementation

[0039] like Figure 1 As shown, the present invention provides a peak-shaving and valley-filling energy storage battery station dispatch and control system, including a cloud dispatch center, an edge control terminal and an execution terminal;

[0040] The execution terminal includes a monitoring module, an energy storage battery monitoring module, a data transmission module, and an energy storage battery control module. The monitoring module monitors power fluctuations in the distributed energy storage cluster to determine peak shaving and valley filling reference points. The energy storage battery monitoring module monitors the dynamic changes in vanadium ion concentrations at the positive and negative electrodes in the battery electrolyte during charging and discharging, as well as the operating rate of the liquid pump under these dynamic changes. The data transmission module transmits power fluctuation data, dynamic changes in vanadium ion concentrations at different valence states, and the operating rate of the liquid pump under these dynamic changes to the peak shaving and valley filling reference points. The energy storage battery control module receives commands and controls the charging and discharging of the energy storage battery.

[0041] The cloud-based dispatch center includes a data receiving module, a data preprocessing module, and a calculation module. The data receiving module receives power fluctuation data at peak shaving and valley filling reference points, dynamic change image data of vanadium ion concentrations at different valence states, power grid data, and the operating rate of the liquid pump under dynamic changes in vanadium ions, sent by the execution terminal. The data preprocessing module corrects the power fluctuation data at peak shaving and valley filling reference points, the dynamic change image data of vanadium ion concentrations at different valence states, and the operating rate of the liquid pump. The calculation module constructs a comprehensive model based on the corrected power fluctuation data and image data, combined with the power grid transmission data, the dynamic change image data of vanadium ion concentrations at different valence states, and the data corrected by the liquid pump operating rate, and formulates a power allocation plan based on the model.

[0042] The edge control terminal includes a receiving control module, which is used to receive instructions from the cloud dispatch center, allocate power to each energy storage station in the adjustment area, and send execution instructions to the execution terminal during peak shaving and valley filling periods.

[0043] At the same time, when collecting data, relevant data such as environmental and weather data can be combined to build a comprehensive model, thereby increasing the scope of judgment of the dispatch and control system;

[0044] The dynamic changes in the concentrations of vanadium ions in different valence states in the positive and negative states of the battery fluid were analyzed using ultraviolet-visible spectrophotometry. Because the vanadium redox flow battery operates through a pump circulation system during charging and discharging, ensuring continuous flow and reaction of the battery's active materials, the specific reactions of the battery fluid during the entire discharge process include the following:

[0045] Discharge process: Positive electrode: ,negative electrode: ;

[0046] Charging process: Positive electrode: ,negative electrode: ;

[0047] Vanadium ions in different valence states ( , , , Vanadium ions have characteristic absorption peaks in the ultraviolet-visible region. By measuring the absorbance of the electrolyte at a specific wavelength, the concentration of vanadium ions in different valence states can be quantitatively analyzed, specifically through ultraviolet or visible light radiation instruments.

[0048] The power fluctuation data and pump operating rate data at the peak shaving and valley filling reference points are combined with the time series data to form an LSTM model St= Make corrections;

[0049] Furthermore, an LSTM model is constructed to learn the patterns of time series data during peak-shaving and valley-filling periods and predict normal values. It takes historical time window data (e.g., the previous 10 time steps) as input and outputs the predicted value for the current time step. The training objective is to minimize the mean squared error between the predicted and actual values. The trained LSTM model is used to predict for each time step, and the residual between the actual and predicted values ​​is calculated.

[0050] Residual = |ytrue−ypredicted|, set a threshold (e.g., residual exceeds 3 times the standard deviation), mark outliers, replace outliers with predicted values ​​and perform linear interpolation or weighted smoothing by combining the normal values ​​before and after, and re-input the corrected values ​​into the model;

[0051] Specifically, in the aforementioned model revision, S t The hidden state at the current time step t; The input vector at the current time step t; The weight matrix is ​​input into the hidden state; This is the hidden state of the previous time step t-1; Let f be the weight matrix from hidden state to hidden state, and f be the activation function;

[0052] Image data correction based on dynamic changes in vanadium ion concentration at different valence states is achieved through a combination of CNN loss function and model training. Specifically, the CNN architecture is designed, including convolutional layers, activation functions (such as ReLU), pooling layers, and fully connected layers. The model's layers are designed as follows: Convolutional layers (Conv2D), activation layers (Activation), pooling layers (MaxPooling), batch normalization layers (Batch Normalization), Dropout layers (to prevent overfitting), and fully connected layers (Dense). Simultaneously, the shape of the input layer should match the shape of the dynamic image data, and the shape of the output layer should match the shape of the target image. The CNN loss function is as follows:

[0053] ;

[0054] in, The loss function; For this is the first The true label of each sample; For this is the first The predicted probability of each sample;

[0055] In addition to the above-mentioned convolutional neural network (CNN) method, model training can also be performed based on gated recurrent unit (GRU), temporal convolutional network (TCN) and other model training methods.

[0056] The modified model data of different categories are evolved into a comprehensive model. The comprehensive model is generated based on the power fluctuation data of the peak shaving period reference point and the valley filling period reference point, the dynamic change image data model of vanadium ion concentration of different valence states, and the operating rate model of the liquid pump.

[0057] The models for power fluctuation data at peak-shaving and valley-filling time reference points use the peak-shaving time reference point as the starting point and the valley-filling time reference point as the ending point. The models are constructed by combining the changes in charging and discharging power of the energy storage battery with battery loss costs throughout the entire time period. The objective function is as follows:

[0058] Where Pbat is the energy storage charging and discharging power (positive for discharging, negative for charging); ΔPref is the deviation between the reference point and the current load; and Dbat is the battery loss cost.

[0059] Furthermore, the power fluctuation data at the peak shaving period benchmark and the valley filling period benchmark are the changes in power load at the peak shaving period benchmark and the valley filling period benchmark. The changes in load are the changes generated during the charging and discharging of the energy storage battery. The battery loss cost is calculated by combining the power flow data, environmental temperature and humidity data and other relevant estimated data during each charging and discharging of the energy storage battery. The model constructed by the fluctuation data and the evolution of battery loss cost can realize the monitoring of energy storage battery charging and discharging and the determination of battery life and usage.

[0060] Meanwhile, the power fluctuation data for peak shaving and valley filling reference points need to be constrained by time and charging / discharging states. Based on the power fluctuation data of peak shaving and valley filling reference points, a constraint model is constructed, and its constraint formulas include the following: ,in In time The state of the battery at that time (charging, discharging); In time The state of the battery at that time (charging, discharging);

[0061] For charge and discharge efficiency; This refers to the battery's rated capacity. In time Battery power at that time; For time step;

[0062] The image data model for the dynamic changes in vanadium ion concentration at different valence states is based on the conservation of momentum of vanadium ions in the battery electrolyte during charging and discharging. and Combining the concentration changes with the changes in vanadium ions in the battery solution during the charging and discharging of the vanadium redox flow battery, it can be concluded that during charging, Oxidized In the positive electrode electrolyte Concentration increases Concentration decreases; during discharge, Restored to , Concentration decreases, As the concentration increases, based on the changes in vanadium ion concentration in the battery solution, dynamic image features are extracted, and the Lucas-Kanades algorithm combined with a mathematical model is used to calculate the formula. Build a model;

[0063] The operating rate of the liquid pump is the core factor affecting the change in vanadium ion concentration. This model is based on flow characteristics and pump dynamics, and is constructed using a first-order linear differential equation. The specific formula is as follows:

[0064] ;

[0065] in, This is the pump's gain coefficient; This refers to the pump's input pressure. This refers to the pump's output pressure.

[0066] Specifically, by modeling the changes in vanadium ion concentration in different valence states in the battery fluid and the evolution of the pump flow, the true state of the energy storage battery during charging and discharging can be directly reflected. Through the secondary evolution of this model and the power fluctuation data model during the charging and discharging of the energy storage battery, a model for determining the power flow of the energy storage battery can be generated, which improves the accuracy of the energy storage battery flow data and enables effective determination of the battery state.

[0067] Regarding the allocation and management methods for peak-shaving and valley-filling energy storage battery stations, the specific methods include the following:

[0068] Collect power grid data and charging and discharging data of each energy storage battery in the distributed energy storage cluster to determine the benchmark points for peak shaving and valley filling periods. Collect fluctuation data of each energy storage battery based on power load within the benchmark points for peak shaving and valley filling periods. After correction, upload the data to the cloud dispatch center. Combine the battery loss cost data to build a model. Through model evolution, generate model A.

[0069] During peak shaving and valley filling periods, data were collected at reference points for each energy storage battery during charging and discharging, including the concentrations of electrolyte in the positive and negative electrodes. and The dynamically changing image data and the data on the operating rate of the liquid pump during the charging and discharging of the energy storage battery are corrected and uploaded to the cloud dispatch center to build a model. Through model evolution, model B is generated.

[0070] Model A and Model B are evolved to generate Model C. Based on the power load during peak shaving and valley filling periods, Model C can predict the charging and discharging of each energy storage battery in the energy storage cluster and send power allocation instructions to the edge control terminal. During peak shaving and valley filling periods, execution instructions are sent to the execution terminal to precisely control the charging and discharging of each energy storage battery in the energy storage cluster.

[0071] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Those skilled in the art can readily implement the invention based on the accompanying drawings and the description above. However, any modifications, alterations, or variations made by those skilled in the art without departing from the scope of the present invention, using the disclosed technical content, are equivalent embodiments of the present invention. Furthermore, any modifications, alterations, or variations made to the above embodiments based on the essential technology of the present invention are still within the protection scope of the present invention.

Claims

1. A peak-shaving and valley-filling energy storage battery station dispatching and control system, characterized in that, include: Cloud-based dispatch center, edge control terminal, and execution terminal; The execution terminal includes a monitoring module, an energy storage battery monitoring module, a data transmission module, and an energy storage battery control module; The system includes several modules: a monitoring module to monitor power fluctuations in the distributed energy storage cluster and determine peak-shaving and valley-filling reference points; a battery monitoring module to monitor the dynamic changes in vanadium ion concentrations at the positive and negative electrodes during charging and discharging, as well as the pump operating rate under these dynamic changes; a data transmission module to transmit power fluctuation data, dynamic vanadium ion concentration image data, and pump operating rate under these dynamic changes; and a battery control module to receive commands and control the charging and discharging of the energy storage battery. The cloud-based dispatch center includes a data receiving module, a data preprocessing module, and a computing module; The data receiving module receives power fluctuation data at peak shaving and valley filling reference points, dynamic change image data of vanadium ion concentrations at different valence states, power grid data, and the operating rate of the liquid pump under dynamic changes in vanadium ions from the execution terminal. The data preprocessing module corrects the power fluctuation data at peak shaving and valley filling reference points, the dynamic change image data of vanadium ion concentrations at different valence states, and the operating rate of the liquid pump. The calculation module constructs a comprehensive model based on the corrected power fluctuation data and image data, combined with the power grid transmission data, the dynamic change image data of vanadium ion concentrations at different valence states, and the data corrected by the liquid pump operating rate, and formulates a power allocation plan based on the model. The edge control terminal includes a receiving control module, which receives instructions from the cloud dispatch center, distributes power to each energy storage station within the adjustment area, and sends execution instructions to the execution terminal during peak shaving and valley filling periods.

2. The peak-shaving and valley-filling energy storage battery station dispatching and control system according to claim 1, characterized in that: The dynamic changes in the concentration of vanadium ions in different valence states at the positive and negative electrodes in the battery solution were investigated using ultraviolet-visible spectrophotometry.

3. The peak-shaving and valley-filling energy storage battery station dispatching and control system according to claim 1, characterized in that: The power fluctuation data and pump operating rate data at the peak shaving and valley filling reference points are combined with the time series data to form an LSTM model St= Make corrections; Where St is the hidden state at the current time step t; The input vector at the current time step t; The weight matrix is ​​input into the hidden state; This is the hidden state of the previous time step t-1; is the weight matrix from hidden state to hidden state; f is the activation function.

4. The peak-shaving and valley-filling energy storage battery station dispatching and control system according to claim 1, characterized in that: The dynamic changes in vanadium ion concentrations at different valence states were corrected using image data, which was achieved through a combination of CNN loss function and model training. The loss function is as follows: ； in, The loss function; For this is the first The true label of each sample; For this is the first The predicted probability of a sample.

5. The peak-shaving and valley-filling energy storage battery station dispatching and control system according to claim 1, characterized in that: The integrated model is generated by evolving from power grid transmission data, power fluctuation data at the reference points of peak shaving and valley filling periods, dynamic change image data model of vanadium ion concentration in different valence states, and liquid pump operating rate model. The models for power fluctuation data at peak-shaving and valley-filling time reference points use the peak-shaving time reference point as the starting point and the valley-filling time reference point as the ending point. The models are constructed by combining the changes in charging and discharging power of the energy storage battery with battery loss costs throughout the entire time period. The objective function is as follows: Where Pbat is the energy storage charging and discharging power (positive for discharging, negative for charging); ΔPref is the deviation between the reference point and the current load; and Dbat is the battery loss cost.

6. The peak-shaving and valley-filling energy storage battery station dispatching and control system according to claim 5, characterized in that: The dynamic change image data model of vanadium ion concentration in different valence states is based on the conservation of momentum of vanadium ions in the energy storage battery fluid during charging and discharging. and Concentration changes; That is, when charging, Oxidized In the positive electrode electrolyte Concentration increases Concentration decreases; during discharge, Restored to , Concentration decreases, As concentration increases, dynamic image features are extracted, and the Lucas-Kanades algorithm combined with a mathematical model is used to calculate the formula: Build.

7. The peak-shaving and valley-filling energy storage battery station dispatching and control system according to claim 5, characterized in that: The operating rate model of the liquid pump is based on the flow characteristics and pump dynamics, and is constructed using a first-order linear differential equation. The specific formula is as follows: ； in, This is the pump's gain coefficient; This refers to the pump's input pressure. This refers to the pump's output pressure.

8. The peak-shaving and valley-filling energy storage battery station dispatching and control system according to claim 5, characterized in that: The constraints for the model of power fluctuation data at the reference points for peak shaving and valley filling periods include the following: ,in In order to be in The state of the battery at time (charging, discharging); In time The state of the battery at that time (charging, discharging); For charge and discharge efficiency; This refers to the battery's rated capacity. In time Battery power at that time; For time step.

9. A peak-shaving and valley-filling energy storage battery station dispatching and control method according to any one of claims 1-7, characterized in that: Specifically, it includes the following: Collect power grid data and charging and discharging data of each energy storage battery in the distributed energy storage cluster to determine the benchmark points for peak shaving and valley filling periods. Collect fluctuation data of each energy storage battery based on power load within the benchmark points for peak shaving and valley filling periods. After correction, upload the data to the cloud dispatch center. Combine the battery loss cost data to build a model. Through model evolution, generate model A. During peak shaving and valley filling periods, data were collected at reference points for each energy storage battery during charging and discharging, including the concentrations of electrolyte in the positive and negative electrodes. and The dynamically changing image data and the data on the operating rate of the liquid pump during the charging and discharging of the energy storage battery are corrected and uploaded to the cloud dispatch center to build a model. Through model evolution, model B is generated. Model A and Model B are evolved to generate Model C. Based on the power load during peak shaving and valley filling periods, Model C can predict the charging and discharging of each energy storage battery in the energy storage cluster and send power allocation instructions to the edge control terminal. During peak shaving and valley filling periods, execution instructions are sent to the execution terminal to precisely control the charging and discharging of each energy storage battery in the energy storage cluster.

Images