A method for predicting the charge and discharge capacity of an energy storage power station

Abstract

The application discloses a kind of energy storage power station charge-discharge capacity prediction method, it is related to energy storage power station technical field, including: collecting a variety of data in the operation process of energy storage power station, and the original data collected is preprocessed;Based on the theory of deep learning, a multi-factor coupled neural network model is constructed, and the neural network model is trained using the preprocessed data, and the neural network model is used to output the charge-discharge capacity prediction value according to the input data;Real-time data collected and preprocessed are input into the trained neural network model, to obtain the charge-discharge capacity prediction value of the current time energy storage power station;Establish a feedback mechanism, compare the actual charge-discharge capacity measurement value with the prediction value to calculate the prediction error, and when the prediction error meets the set conditions, the neural network model is updated using new data.

Description

Technical Field

[0001] This invention relates to the field of energy storage power station technology, and in particular to a method for predicting the charging and discharging capacity of an energy storage power station. Background Technology

[0002] Against the backdrop of the rapid development of new energy power systems, energy storage power stations, as core equipment for smoothing out fluctuations in renewable energy output and ensuring the stable operation of the power grid, directly affect the dispatch efficiency and energy utilization benefits of the power system. Charging and discharging capacity is a key operating indicator for energy storage power stations. Accurate prediction of charging and discharging capacity can provide reliable decision-making basis for power grid dispatching departments, help to optimize the allocation of energy storage resources, and reduce operating costs.

[0003] Currently, various methods for predicting the charging and discharging capabilities of energy storage power stations have been proposed. For example, patent document CN113612245A discloses a prediction method that identifies dates of the same type as the prediction target, acquires data at the corresponding time points, performs averaging, obtains the predicted charging / discharging capacity, and then corrects it based on the rated power. This method improves the accuracy of the prediction to some extent.

[0004] However, the aforementioned methods largely rely on statistical analysis of historical data and fail to fully consider the complex coupling relationships between multiple factors such as battery status, environmental parameters, and grid load demand on charge and discharge capabilities. When the operating environment of the energy storage power station changes or battery performance degrades, the prediction accuracy of such methods is difficult to guarantee, and they cannot effectively adapt to complex actual operating scenarios.

[0005] To address this, the present invention provides a method for predicting the charging and discharging capacity of an energy storage power station, aiming to solve the problems of insufficient prediction accuracy and difficulty in adapting to complex operating scenarios caused by ignoring the coupling relationship of multiple factors in the prior art. Summary of the Invention

[0006] The main objective of this invention is to provide a method for predicting the charging and discharging capacity of an energy storage power station.

[0007] Another objective of this invention is to provide a device for predicting the charging and discharging capacity of an energy storage power station.

[0008] The third objective of this invention is to provide a computer device.

[0009] A fourth objective of this invention is to provide a non-transitory computer-readable storage medium.

[0010] To achieve the above objectives, a first aspect of the present invention provides a method for predicting the charging and discharging capacity of an energy storage power station, comprising:

[0011] Collect various data during the operation of the energy storage power station and preprocess the collected raw data; A multi-factor coupled neural network model is constructed based on deep learning theory. The neural network model is trained using preprocessed data. The neural network model is used to output a predicted value of charging and discharging capability based on the input data. The real-time collected and preprocessed data is input into the trained neural network model to obtain the predicted value of the charging and discharging capacity of the energy storage power station at the current moment. A feedback mechanism is established to compare the actual charge / discharge capacity measurement value with the predicted value to calculate the prediction error. When the prediction error meets the set conditions, the neural network model is updated using new data.

[0012] In one embodiment of the present invention, the process of collecting various data during the operation of the energy storage power station and preprocessing the collected raw data includes: The various data include battery status data, environmental parameter data, grid load demand data, and charge / discharge strategy data. The battery status data includes SOC and SOH, and the environmental parameter data includes temperature and humidity. The preprocessing includes data cleaning, smoothing filtering, and normalization.

[0013] In one embodiment of the present invention, the step of constructing a multi-factor coupled neural network model based on deep learning theory and training the neural network model using preprocessed data includes: When training the neural network model, the backpropagation algorithm and optimization algorithm are used to adjust the model parameters, the validation set is used to monitor the model performance, and the test set is used to evaluate the model's generalization ability. The neural network model adopts a multilayer perceptron network structure or a long short-term memory network structure, and the optimization algorithm is stochastic gradient descent or Adam algorithm.

[0014] In one embodiment of the present invention, the step of inputting the real-time collected and preprocessed data into a trained neural network model to obtain a predicted value of the charging and discharging capacity of the energy storage power station at the current moment includes: The real-time collected and preprocessed data is organized according to the input layer format and then input into the trained neural network model. The model calculates and outputs the predicted value of the charging and discharging capacity of the energy storage power station at the current moment through forward propagation. The predicted value includes charging capacity and discharging capacity.

[0015] In one embodiment of the present invention, establishing a feedback mechanism involves comparing the actual charge / discharge capacity measurement value with the predicted value to calculate the prediction error. When the prediction error meets a set condition, the neural network model is updated using new data, including: The prediction error is calculated by comparing the actual charge / discharge capacity measurement with the predicted value. The prediction error calculation formula is as follows:

[0016] in, This represents the actual charge / discharge capacity value. These are the model's predicted values.

[0017] Determine whether the prediction error meets the set conditions, the set conditions including the prediction error exceeding a set threshold, or the prediction error exceeding a set threshold for a consecutive number of times; When the prediction error meets the set conditions, new data is added to the training set for incremental training using an online learning method to update or optimize the structure of the neural network model.

[0018] To achieve the above objectives, a second aspect of the present invention provides a device for predicting the charging and discharging capacity of an energy storage power station, comprising: The data acquisition module is used to collect various data during the operation of the energy storage power station and to preprocess the collected raw data. The model building module is used to build a multi-factor coupled neural network model based on deep learning theory, and to train the neural network model using preprocessed data. The neural network model is used to output a predicted value of charging and discharging capability based on the input data. The prediction module is used to input the real-time collected and preprocessed data into the trained neural network model to obtain the predicted value of the charging and discharging capacity of the energy storage power station at the current moment. The correction and optimization module is used to establish a feedback mechanism, compare the actual charge and discharge capacity measurement value with the predicted value to calculate the prediction error, and update the neural network model with new data when the prediction error meets the set conditions.

[0019] To achieve the above objectives, a third aspect of this application provides a computer device comprising a processor and a memory; wherein the processor reads executable program code stored in the memory to run a program corresponding to the executable program code, for implementing a method for predicting the charging and discharging capacity of an energy storage power station as described in the first aspect embodiment.

[0020] To achieve the above objectives, a fourth aspect of this application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for predicting the charging and discharging capacity of an energy storage power station as described in the first aspect embodiment.

[0021] The embodiments of the present invention have the following beneficial effects: 1. This invention constructs a multi-factor coupled neural network model based on deep learning theory. Through multi-layer nonlinear transformation of the hidden layer, it fully explores the complex coupling relationship between multiple factors such as battery state, environmental parameters, and grid load, overcoming the problem of insufficient accuracy caused by neglecting the coupling effect of multiple factors in existing prediction methods. At the same time, it is equipped with a full-process preprocessing scheme of "data cleaning-smoothing filtering-normalization" to ensure the quality of input data, greatly reduce prediction deviation, and significantly improve the prediction accuracy of charge and discharge capacity.

[0022] 2. This method establishes a prediction result feedback and online learning adjustment mechanism. When the prediction error exceeds the threshold, incremental training is used to update the model parameters or optimize the model structure without retraining the entire model. It can dynamically adapt to scenarios such as battery performance degradation and changes in the operating environment. At the same time, it supports multiple model structures such as multilayer perceptron and long short-term memory network, which can be adapted to the needs of different types of energy storage power stations and has a wide range of application scenarios.

[0023] 3. This method uses optimization algorithms such as Adam to improve model training efficiency. In the real-time prediction stage, only data preprocessing and model forward propagation need to be completed, and the computation time is short. Online learning adopts incremental training to further shorten the update time, ensuring the real-time nature of prediction and model adjustment, accurately matching the timeliness requirements of real-time power system dispatch, and providing reliable support for the optimal allocation of energy storage resources. Attached Figure Description

[0024] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 A flowchart illustrating a method for predicting the charging and discharging capacity of an energy storage power station, provided as an embodiment of the present invention; Figure 2 This is a structural diagram of a device for predicting the charging and discharging capacity of an energy storage power station, provided in an embodiment of the present invention. Detailed Implementation

[0025] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] The following describes a method for predicting the charging and discharging capacity of an energy storage power station according to an embodiment of the present invention, with reference to the accompanying drawings.

[0028] Example 1 This embodiment provides a method for predicting the charging and discharging capacity of an energy storage power station. For example... Figure 1 As shown, the method includes the following steps: S1 collects various data during the operation of the energy storage power station and preprocesses the collected raw data.

[0029] Specifically, the collected data includes battery status data, environmental parameter data, grid load demand data, and charge / discharge strategy data. Battery status data includes SOC (State of Charge) and SOH (State of Health), while environmental parameter data includes temperature and humidity. The specific data ranges are as follows: battery status data: SOC 20%–80%, SOH 90%–98%; environmental parameter data: temperature -5–35℃, humidity 30%–80%; grid load demand data: 100–500kW; charge / discharge strategy data: charge / discharge rate 0.2C–1C, charge / discharge duration 0.5–4h.

[0030] Furthermore, data acquisition is achieved by installing corresponding sensors at the battery packs, environmental monitoring points, and grid connection points of the energy storage power station. In this embodiment, voltage sensors, current sensors, temperature sensors, humidity sensors, and power sensors are installed at the battery packs, environmental monitoring points, and grid connection points of the energy storage power station to collect continuous operational data for 30 days in real time, with a collection frequency of once every 5 minutes.

[0031] The collected raw data was cleaned, and outliers were removed using the 3σ criterion. A data smoothing filter algorithm was used to smooth the data and remove noise interference. Finally, the maximum-minimum normalization method was used to map all data to the [0,1] interval. The normalization formula is as follows:

[0032] in, The original data, and These are the minimum and maximum values ​​of the data, respectively. This is the normalized data.

[0033] Based on deep learning theory, a multi-factor coupled neural network model is constructed. Through multi-layer nonlinear transformation of the hidden layer, it can fully explore the complex coupling relationship between multiple factors such as battery state, environmental parameters, and grid load, which solves the shortcomings of existing technologies that analyze single factors in isolation. At the same time, it is equipped with a full-process preprocessing scheme of "data cleaning-smoothing filtering-normalization" to ensure the quality of input data, significantly reduce prediction bias, and significantly improve the prediction accuracy of charge and discharge capacity.

[0034] S2, Construct a multi-factor coupled neural network model based on deep learning theory, and train the neural network model using preprocessed data. The neural network model is used to output a predicted value of charging and discharging capability based on the input data.

[0035] In this embodiment, the constructed multi-factor coupled neural network model adopts a multilayer perceptron or long short-term memory network structure. Taking a long short-term memory network (LSTM) as an example, the number of nodes in the input layer is determined according to the data type and dimension of the collected data; multiple hidden layers are set, and the number of nodes in each layer is set according to the actual situation; the number of nodes in the output layer is 1, which is used to output the predicted values ​​of charging capacity and discharging capacity respectively.

[0036] Furthermore, operational data from energy storage power stations over a past period are collected as training samples and divided into training, validation, and test sets according to a certain ratio. When training the model using the training set, backpropagation and optimization algorithms are employed to adjust model parameters. The optimization algorithm is either stochastic gradient descent or the Adam algorithm. In this embodiment, the Adam optimization algorithm is used, with the learning rate and batch size set. During training, the model performance is evaluated using the validation set at regular intervals. Training stops when the validation set loss function value no longer decreases after a certain number of consecutive iterations to prevent overfitting. After training, the generalization ability of the model is evaluated using the test set.

[0037] S3 inputs the real-time collected and preprocessed data into the trained neural network model to obtain the predicted value of the charging and discharging capacity of the energy storage power station at the current moment.

[0038] Specifically, the real-time collected and preprocessed data is organized according to the input layer format and then input into the trained neural network model. The model calculates and outputs the predicted value of the charging and discharging capacity of the energy storage power station at the current moment through forward propagation. The predicted value includes charging capacity and discharging capacity.

[0039] S4. Establish a feedback mechanism, compare the actual charge / discharge capacity measurement value with the predicted value to calculate the prediction error, and when the prediction error meets the set conditions, update the neural network model using new data.

[0040] In this embodiment, when the feedback mechanism is established, the actual measured charging and discharging capacity value of the energy storage power station is compared and analyzed with the model prediction value at regular intervals (e.g., 1 hour). The prediction error is calculated using the following formula:

[0041] in, This represents the actual charge / discharge capacity value. These are the model's predicted values.

[0042] In this embodiment, the prediction error is determined to meet a set threshold of 5%. If the prediction error exceeds 5% for a set number of consecutive times (e.g., 3 times), a dynamic model adjustment mechanism is activated. New real-time data and actual measurements are added to the training set using online learning to incrementally train the model, thereby updating or optimizing the structure of the neural network model. Simultaneously, based on the analysis results of the prediction error, the structure of the LSTM model can be optimized, such as by increasing or decreasing the number of hidden layer nodes or adjusting the number of layers. The model is then retrained until the prediction error meets the requirements.

[0043] Specifically, by establishing a sound prediction result feedback and online learning adjustment mechanism, when the prediction error exceeds the threshold, the model parameters can be updated and the structure optimized through incremental training without retraining the entire model. This allows for dynamic adaptation to scenarios such as battery performance degradation and changes in the operating environment. At the same time, it supports multiple model structures such as multilayer perceptrons and long short-term memory networks, which can be adapted to the needs of different types of energy storage power stations and have a wide range of application scenarios.

[0044] By optimizing the algorithm to improve model training efficiency, the real-time prediction stage only requires data preprocessing and model forward propagation, resulting in short computation time. Online learning adopts incremental training to further shorten update time, ensuring the real-time nature of prediction and model adjustment. This can accurately match the timeliness requirements of real-time power system dispatch and provide reliable support for the optimal allocation of energy storage resources.

[0045] Example 2 This invention also provides a device for predicting the charging and discharging capacity of an energy storage power station, such as... Figure 2 As shown, the device 10 includes: The data acquisition module 100 is used to collect various data during the operation of the energy storage power station and to preprocess the collected raw data. The model building module 200 is used to build a multi-factor coupled neural network model based on deep learning theory, and to train the neural network model using preprocessed data. The neural network model is used to output a predicted value of charging and discharging capacity based on the input data. The prediction module 300 is used to input the real-time collected and preprocessed data into the trained neural network model to obtain the predicted value of the charging and discharging capacity of the energy storage power station at the current moment. The correction and optimization module 400 is used to establish a feedback mechanism, compare the actual charge and discharge capacity measurement value with the predicted value to calculate the prediction error, and update the neural network model with new data when the prediction error meets the set conditions.

[0046] Example 3 To implement the methods of the above embodiments, the present invention also provides a computer device, which includes a memory and a processor; wherein the processor runs a program corresponding to the executable program code by reading executable program code stored in the memory, so as to implement the various steps of the methods described above.

[0047] Example 4 To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method described in the foregoing embodiments.

[0048] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0049] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0050] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A method for predicting the charge and discharge capacity of an energy storage power station, characterized in that, Includes the following steps: Collect various data during the operation of the energy storage power station and preprocess the collected raw data; A multi-factor coupled neural network model is constructed based on deep learning theory. The neural network model is trained using preprocessed data. The neural network model is used to output a predicted value of charging and discharging capability based on the input data. The real-time collected and preprocessed data is input into the trained neural network model to obtain the predicted value of the charging and discharging capacity of the energy storage power station at the current moment. A feedback mechanism is established to compare the actual charge / discharge capacity measurement value with the predicted value to calculate the prediction error. When the prediction error meets the set conditions, the neural network model is updated using new data.

2. The method according to claim 1, characterized in that, The process involves collecting various data during the operation of the energy storage power station and preprocessing the collected raw data, including: The various data include battery status data, environmental parameter data, grid load demand data, and charge / discharge strategy data. The battery status data includes SOC and SOH, and the environmental parameter data includes temperature and humidity. The preprocessing includes data cleaning, smoothing filtering, and normalization.

3. The method of claim 1, wherein, The method for constructing a multi-factor coupled neural network model based on deep learning theory, and training the neural network model using preprocessed data, includes: When training the neural network model, the backpropagation algorithm and optimization algorithm are used to adjust the model parameters, the validation set is used to monitor the model performance, and the test set is used to evaluate the model's generalization ability. The neural network model adopts a multilayer perceptron network structure or a long short-term memory network structure, and the optimization algorithm is stochastic gradient descent or Adam algorithm.

4. The method of claim 1, wherein, The step of inputting the real-time collected and preprocessed data into the trained neural network model to obtain the predicted value of the charging and discharging capacity of the energy storage power station at the current moment includes: The real-time collected and preprocessed data is organized according to the input layer format and then input into the trained neural network model. The model calculates and outputs the predicted value of the charging and discharging capacity of the energy storage power station at the current moment through forward propagation. The predicted value includes charging capacity and discharging capacity.

5. The method of claim 1, wherein, The establishment of a feedback mechanism involves comparing the actual charge / discharge capacity measurement with the predicted value to calculate the prediction error. When the prediction error meets a set condition, the neural network model is updated using new data, including: The prediction error is calculated by comparing the actual charge / discharge capacity measurement with the predicted value. The prediction error calculation formula is as follows: wherein, is the actual charge and discharge capacity value, is the model prediction value. Determine whether the prediction error meets the set conditions, the set conditions including the prediction error exceeding a set threshold, or the prediction error exceeding a set threshold for a consecutive number of times; When the prediction error meets the set conditions, new data is added to the training set for incremental training using an online learning method to update or optimize the structure of the neural network model.

6. A device for predicting the charge and discharge capacity of an energy storage power station, characterized in that, include: The data acquisition module is used to collect various data during the operation of the energy storage power station and to preprocess the collected raw data. The model building module is used to build a multi-factor coupled neural network model based on deep learning theory, and to train the neural network model using preprocessed data. The neural network model is used to output a predicted value of charging and discharging capability based on the input data. The prediction module is used to input the real-time collected and preprocessed data into the trained neural network model to obtain the predicted value of the charging and discharging capacity of the energy storage power station at the current moment. The correction and optimization module is used to establish a feedback mechanism, compare the actual charge and discharge capacity measurement value with the predicted value to calculate the prediction error, and update the neural network model with new data when the prediction error meets the set conditions.

7. A computer device, comprising: Including processor and memory; The processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement a method for predicting the charging and discharging capacity of an energy storage power station as described in any one of claims 1-5.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, When executed by the processor, the program implements a method for predicting the charging and discharging capacity of an energy storage power station as described in any one of claims 1-5.

Images