Energy storage system initial capacity prediction method and device, equipment and storage medium

Abstract

The application provides a method and device for predicting an initial capacity of an energy storage system, equipment and a storage medium. The method comprises: obtaining parameter values of multiple parameter types of multiple battery cells of the energy storage system; the multiple parameter types include multiple types from the following types: battery cell capacity, battery cell voltage, battery cell temperature and battery cell resistance; based on the parameter values of the multiple parameter types, a first input feature map corresponding to each of the multiple parameter types is constructed; the parameter types of the parameter values in the first input feature map are the same; the constructed first input feature map is input into a pre-trained AI model to predict an initial capacity of the energy storage system and determine a predicted value of the initial capacity.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of batteries, and relates to, but is not limited to, an initial capacity prediction method and device for an energy storage system, equipment and a storage medium. BACKGROUND

[0002] The energy storage system integrated by multiple battery cells refers to combining multiple single battery cells in a way of series connection, parallel connection or mixed connection to form a complete battery module or battery pack. This integration manner can improve the voltage, capacity and power of the battery system to meet the needs of different application scenarios. SUMMARY

[0003] The present application provides an initial capacity prediction method and device for an energy storage system, equipment and a storage medium, wherein:

[0004] In a first aspect, the present application provides an initial capacity prediction method for an energy storage system, which comprises: obtaining parameter values of multiple parameter types of multiple battery cells of the energy storage system; the multiple parameter types include multiple items in the following types: battery cell capacity, battery cell voltage, battery cell temperature and battery cell resistance; based on the parameter values of the multiple parameter types, a first input feature map corresponding to each of the multiple parameter types is constructed; wherein the parameter values in the first input feature map are of the same parameter type; and the constructed first input feature map is input into a pre-trained AI model to predict the initial capacity of the energy storage system and determine a predicted value of the initial capacity.

[0005] It can be understood that in the present application, a method capable of predicting the initial capacity of an energy storage system before multiple battery cells are integrated into the energy storage system is provided, which fills the technical gap in predicting the initial capacity of the energy storage system. The method predicts a predicted value of the initial capacity of the energy storage system based on a first input feature map of multiple parameter values of the same parameter type of the multiple battery cells. In this way, without integrating the multiple battery cells into the energy storage system, the initial capacity of the energy storage system is measured and obtained, thereby benefiting the manufacturer of the energy storage system to determine the number of battery cells of the to-be-integrated energy storage system for reference based on the predicted value with high accuracy, and further benefiting the reduction of the hardware cost of the energy storage system while avoiding complaints due to the initial capacity not meeting the requirements of customers.

[0006] It can be understood that in the present application, the inventors of the present application find that when multiple battery cells are integrated into an energy storage system, the battery cell capacity, battery cell voltage, battery cell temperature or battery cell resistance of the multiple battery cells will interact with each other and thus affect the initial capacity of the energy storage system. Therefore, in the present application, the initial capacity of the energy storage system is determined based on the first input feature map including the above-mentioned parameter types, which is beneficial to improving the accuracy of the predicted value of the predicted initial capacity by determining the key factors affecting the initial capacity of the energy storage system.

[0007] In some embodiments, the energy storage system includes a first number of battery cells; the first input feature map includes parameter values of a same parameter type of the first number of battery cells; and the parameter values in the first input feature map are arranged in order of sizes of battery cell capacities of the first number of battery cells.

[0008] It can be understood that, in the embodiments of the present application, the parameter values in the first input feature map are arranged in order of sizes of battery cell capacities of the first number of battery cells because the inventors of the present application have found through experiments that arranging the parameter values in the first input feature map in order of sizes of battery cell capacities can improve the accuracy of the predicted value of the initial capacity of the energy storage system, thereby benefiting the manufacturer of the energy storage system to determine the number of battery cells required for the to-be-integrated energy storage system based on the predicted value with high accuracy, and further benefiting to ensure that the initial capacity of the energy storage system integrated based on a plurality of battery cells meets the requirements of customers while reducing the hardware cost of the energy storage system.

[0009] In some embodiments, the method further includes: inputting the constructed first input feature map into the pre-trained AI model to predict the initial capacity of the energy storage system and determine the predicted value of the initial capacity.

[0010] It can be understood that, in the embodiments of the present application, the initial capacity of the energy storage system can be predicted based on a plurality of first input feature maps before the plurality of battery cells are integrated into the energy storage system, thereby determining the predicted value of the initial capacity. In this way, the initial capacity of the energy storage system does not need to be obtained after the plurality of battery cells are integrated into the energy storage system, thereby benefiting the manufacturer of the energy storage system to determine the number of battery cells of the to-be-integrated energy storage system based on the predicted value with high accuracy, and further benefiting to reduce the hardware cost of the energy storage system while avoiding complaints due to the initial capacity not meeting the requirements of customers.

[0011] In some embodiments, the method further includes: recommending a target number of battery cells required for integrating the energy storage system based on a size relationship between the predicted value of the initial capacity and a preset expected value.

[0012] It can be understood that, in the embodiments of the present application, the target number of battery cells required for integrating the energy storage system is recommended based on a size relationship between the predicted value of the initial capacity and a preset expected value. In this way, the target number of battery cells required for integrating the energy storage system can be automatically or intelligently recommended, thereby improving the efficiency of determining the target number of battery cells required for integrating the energy storage system and reducing the labor cost of determining the target number of battery cells required for integrating the energy storage system.

[0013] In some embodiments, based on the size relationship between the predicted value of the initial capacity and the preset expected value, the target number of battery cells required for the integrated energy storage system is recommended, including: in the case that the predicted value of the initial capacity meets the first condition, recommending the target number as a first number; wherein the first number is the number of battery cells in the energy storage system; the first condition includes that the predicted value of the initial capacity is greater than the preset expected value, and the difference between the predicted value of the initial capacity and the preset expected value is within a first numerical range.

[0014] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity is greater than the preset expected value and the difference between the predicted value of the initial capacity and the preset expected value is within the first numerical range, the target number is recommended as the first number. That is, when the part of the predicted value of the initial capacity that is higher than the preset expected value is within the acceptable overrun range, the target number is recommended as the first number, so that the recommended target number is more reasonable, and then the initial capacity of the energy storage system integrated based on the reasonable number of battery cells can meet the customer's requirements while reducing the hardware cost of the energy storage system.

[0015] In some embodiments, based on the size relationship between the predicted value of the initial capacity and the preset expected value, the target number of battery cells required for the integrated energy storage system is recommended, including: in the case that the predicted value of the initial capacity does not meet the first condition, determining a plurality of second input feature maps; wherein the second input feature map includes parameter values of a second number of battery cells, and the second number is different from the first number; the parameter values in the second input feature map are of the same parameter type; the same element position of different second input feature maps corresponds to different parameter types of the same battery cell; based on the plurality of second input feature maps and the pre-trained AI model, the predicted value of the initial capacity is re-determined until the predicted value of the initial capacity meets the first condition, and the target number is recommended as the number of battery cells corresponding to the predicted value that meets the first condition.

[0016] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity does not meet the first condition, a plurality of second input feature maps including parameter values of a second number of battery cells are determined, and the second number is different from the first number. That is, in the case that the predicted value of the initial capacity does not meet the first condition, the number of battery cells of the energy storage system to be integrated is updated, the predicted value of the initial capacity is re-predicted until the predicted value of the initial capacity meets the first condition, and then the target number is recommended as the number of battery cells corresponding to the predicted value that meets the first condition. That is, the target number of battery cells required for the integrated energy storage system is recommended only when it is ensured that the part of the predicted value of the initial capacity that is higher than the preset expected value is within the acceptable overrun range. In this way, the recommended target number is more reasonable, that is, the initial capacity of the energy storage system integrated based on the target number of battery cells can meet the customer's requirements while reducing the hardware cost of the energy storage system.

[0017] In some embodiments, the first condition not being satisfied by the predicted value of the initial capacity comprises: the predicted value of the initial capacity being less than or equal to a preset expected value; and the second number being greater than the first number.

[0018] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity does not satisfy or just satisfies the preset expected value, the number of the battery cells of the to-be-integrated energy storage system is increased, and then the predicted value of the initial capacity is re-determined based on the plurality of second input feature maps with the increased number of battery cells and the pre-trained AI model. In this way, it is beneficial to avoid the problem that the energy storage system integrated with the target number of battery cells does not satisfy the requirements of the customer, thereby causing customer complaints.

[0019] In some embodiments, the first condition not being satisfied by the predicted value of the initial capacity comprises: the predicted value of the initial capacity being greater than the preset expected value, and a difference between the predicted value of the initial capacity and the preset expected value being less than a lower limit value of the first numerical range; and the second number being greater than the first number.

[0020] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity is higher than the preset expected value, but the part of the increase is not too much, the number of the battery cells of the to-be-integrated energy storage system is increased, and then the predicted value of the initial capacity is re-determined based on the plurality of second input feature maps with the increased number of battery cells and the pre-trained AI model. In this way, it is beneficial to ensure that the energy storage system integrated with the target number of battery cells satisfies the requirements of the customer, thereby avoiding the problem of customer complaints.

[0021] In some embodiments, the first condition not being satisfied by the predicted value of the initial capacity comprises: the predicted value of the initial capacity being greater than the preset expected value, and a difference between the predicted value of the initial capacity and the preset expected value being greater than an upper limit value of the first numerical range; and the second number being less than the first number.

[0022] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity far exceeds the preset expected value, the number of the battery cells of the to-be-integrated energy storage system is reduced, and then the predicted value of the initial capacity is re-determined based on the plurality of second input feature maps with the reduced number of battery cells and the pre-trained AI model. In this way, it is beneficial to avoid the problem that the energy storage system integrated with the target number of battery cells is far higher than the requirements of the customer, thereby reducing the hardware cost of the energy storage system.

[0023] In a second aspect, an embodiment of the present application provides a device for predicting initial capacity of an energy storage system, the device comprising: an obtaining module configured to obtain parameter values of a plurality of parameter types of a plurality of battery cells of the energy storage system; the plurality of parameter types comprises a plurality of the following types: battery cell capacity, battery cell voltage, battery cell temperature, and battery cell resistance; a constructing module configured to construct a first input feature map corresponding to each of the plurality of parameter types based on the parameter values of the plurality of parameter types; wherein the parameter values in the first input feature map are of the same parameter type; and a predicting module configured to input the constructed first input feature map into a pre-trained AI model, and predict the initial capacity of the energy storage system to determine a predicted value of the initial capacity.

[0024] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory and a processor, the memory storing a computer program capable of running on the processor, and the processor implements the method of the first aspect when executing the program.

[0025] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method of the first aspect.

[0026] In a fifth aspect, an embodiment of the present application provides a computer program product, which comprises a computer program or instructions, and the computer program or instructions are executed by a processor to implement the method of the first aspect.

[0027] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0028] The drawings incorporated in the specification and forming a part thereof illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application. It is clear that the drawings described below are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0029] The flowchart shown in the drawings is only an exemplary description, and is not necessarily required to include all contents and operations / steps, nor is it necessarily required to be executed in the order described. For example, some operations / steps can be further divided, and some operations / steps can be combined or partially combined, so the actual execution order may be changed according to the actual situation.

[0030] Figure 1 An implementation flowchart of a method for predicting initial capacity of an energy storage system provided by an embodiment of the present application;

[0031] Figure 2A schematic diagram of a plurality of first input feature maps provided for an embodiment of the present application;

[0032] Figure 3 A schematic diagram of a first input feature map provided for an embodiment of the present application;

[0033] Figure 4 A schematic diagram of an energy storage system provided for an embodiment of the present application;

[0034] Figure 5 A schematic diagram of a mathematical model of an energy storage system provided for an embodiment of the present application;

[0035] Figure 6 A structural schematic diagram of an AI model provided for an embodiment of the present application;

[0036] Figure 7 A schematic diagram of a convolution operation provided for an embodiment of the present application;

[0037] Figure 8 A schematic diagram of a pooling operation provided for an embodiment of the present application;

[0038] Figure 9 A schematic diagram of an activation function provided for an embodiment of the present application;

[0039] Figure 10 A schematic diagram of a back propagation process of a convolutional neural network provided for an embodiment of the present application;

[0040] Figure 11 An implementation flowchart of a recommended target number of battery cells required by an integrated energy storage system provided for an embodiment of the present application;

[0041] Figure 12 A structural schematic diagram of an energy storage system initial capacity prediction device provided for an embodiment of the present application;

[0042] Figure 13 A structural schematic diagram of an electronic device provided for an embodiment of the present application. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application.

[0044] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described below with reference to the drawings, and the described embodiments should not be regarded as limiting the present application, and all other embodiments obtained by those skilled in the art without making creative efforts fall within the scope of protection of the present application.

[0045] In the following description, "some embodiments / some other embodiments" are described, which describe a subset of all possible embodiments, but it can be understood that "some embodiments / some other embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.

[0046] In the following description, the terms "first / second" are only distinguished from similar objects, and do not represent a specific order of the objects. It can be understood that "first / second" can be interchanged in a specific order or sequence as allowed, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of the application and are not intended to limit the application.

[0048] In related technology one, a seasonal autoregressive moving average model is trained using a training data set, model parameters are updated, a target seasonal autoregressive moving average model is obtained, a current prediction data set is input into the target seasonal autoregressive moving average model, prediction curves of wind power, light power and load are obtained, a load difference value is determined based on the prediction curves of wind power, light power and load, a target prediction error is obtained, and an energy storage capacity prediction value is determined based on the target prediction error and the load difference value.

[0049] In related technology two, a target voltage interval is obtained; a plurality of historical sample data of the battery cells in the target voltage interval are obtained, linear fitting is performed, and a fitting formula is obtained; a first battery cell working voltage and a second battery cell working voltage are selected in the target voltage interval, a difference ΔY between a first battery cell state of charge (SOC) and a second battery cell SOC is obtained according to the fitting formula; discharge capacities of a battery cell to be measured under the first battery cell working voltage and the second battery cell working voltage are obtained, and a battery cell predicted capacity is determined according to a difference ΔC between the first discharge capacity and the second discharge capacity and ΔY.

[0050] However, in related technologies one and two, the prediction of the process capacity of the battery cell is mainly focused on, and there is a lack of relevant technical methods for the capacity of the battery cell integrated cabinet. That is, the related technology only provides a prediction method for the process capacity of a single battery cell, and does not provide a capacity prediction method for the cabinet; wherein the cabinet integrates a plurality of battery cells.

[0051] In other related technologies, process capacity prediction technical solutions are mainly concentrated in the following four aspects:

[0052] (1) Curve fitting prediction: by collecting the capacity value of the initial cycle of the battery, the least squares fitting is performed with an exponential curve, and the correlation coefficient of the curve is optimized to obtain a prediction model;

[0053] (2) Machine learning model prediction: by collecting the historical data of the battery, all relevant parameters of the battery are used as input, and the capacity value is used as target to build a regression model through supervised machine learning. Commonly used models include random forest, eXtreme Gradient Boosting (XGboost), Adaptive Boosting (Adaboost), etc.;

[0054] (3) Neural network model prediction: by collecting the historical data of the battery, all relevant parameters of the battery and the capacity of the battery at the previous time are used as input, and the capacity of the battery at the next time is used as target. Based on the recurrent neural network framework, the long short-term memory (LSTM) is used to optimize the parameters of the neural network hidden layer, thereby building a model for predicting the capacity;

[0055] (4) Mechanism model and data-driven fusion prediction: combine the electrochemical theory to select the factors that affect the battery, collect the relevant data and capacity target value data corresponding to the characteristic factors, and use machine learning model or neural network model for supervised machine learning to build a regression prediction model. In addition, the current latest physical neural network selects the differential equation model in electrochemistry as the loss function for improvement, further improving the regression prediction ability.

[0056] The inventors of the present application have analyzed and researched the above four aspects of related technology and found that the above four aspects of related technology respectively have certain problems; wherein the problems corresponding to the above four aspects of related technology are:

[0057] (1) For curve fitting prediction: this method implies that the prediction model meets the prerequisite of exponential function, and the optimization of model parameters by data is based on this. However, the assumption of exponential function is not reliable, and the prediction accuracy is low;

[0058] (2) For machine learning model prediction: the key factors obtained by model feature engineering will change due to changes in the data set, and the generalization ability and accuracy of the model are in conflict;

[0059] (3) For neural network model prediction: it is difficult to select key features for neural network modeling, and the computational complexity of the model is relatively large;

[0060] (4) For mechanism model and data-driven fusion prediction: when the research on mechanism is not sufficient, the prediction effect of the model cannot meet the expectation.

[0061] The common problems of the related technologies in the above four aspects are:

[0062] (1) They are all technologies for single cell process capacity prediction, and there is no description of multi-cell combination and mutual prediction.

[0063] (2) The prediction of single cell process capacity and the prediction of multi-cell combination initial capacity are completely different problems.

[0064] Therefore, the applicability of the related technologies has a huge bottleneck. The initial capacity test of the energy storage cabinet (i.e., an example of an energy storage system) needs to consider the following problems:

[0065] (1) Test precision: high-precision test equipment and standardized test procedures are needed to ensure the accuracy of the capacity test;

[0066] (2) Dynamic testing: the energy storage cabinet will be affected by various working conditions in actual operation, and needs to be dynamically tested by simulating actual use scenarios;

[0067] (3) Electronic stability system (ESS) pull line needs to test the capacity before the product is offline, the test time is long, the test frequency is high, and the energy is wasted during the test process;

[0068] (4) The low capacity value of the energy storage will cause customer complaints, and the redundant design will increase the hardware cost;

[0069] (5) Engineering practice has proved that the initial capacity of the energy storage is not simply calculated by the capacity of all cells. Since the factors affecting the initial capacity of the energy storage cabinet (i.e., an example of an energy storage system) are complex, the mutual interaction of numerous cells leads to great challenges in initial capacity prediction.

[0070] Based on the above analysis, at present, the initial capacity of the energy storage system integrated by multiple cells cannot be accurately calculated by relying on experience.

[0071] For the determination of the initial capacity of the cabinet or the cabinet, multiple cells are integrated into the cabinet or the cabinet, and a special measuring device is used to measure the initial capacity of the cabinet or the cabinet, so as to obtain the initial capacity of the cabinet or the cabinet.

[0072] However, this method cannot determine the initial capacity of the electric cabinet or the electric box before the plurality of battery cells are integrated into the electric cabinet or the electric box. Therefore, in order to meet the customer's requirement for the initial capacity of the electric cabinet or the electric box, more battery cells are usually integrated into the electric cabinet or the electric box, which leads to an increase in hardware cost. However, if less battery cells are integrated into the electric cabinet or the electric box in order to reduce the hardware cost, more battery cells cannot be added to the electric cabinet or the electric box due to the fixed size of the electric cabinet or the electric box, which causes the problem of customer complaints due to the initial capacity of the electric cabinet or the electric box not meeting the customer's requirement.

[0073] Based on this, the embodiment of the present application provides a method for predicting the initial capacity of an energy storage system, Figure 1 The implementation process schematic diagram of the method for predicting the initial capacity of the energy storage system provided by the embodiment of the present application is shown in Figure 1 As shown in the figure, the method can include the following steps 101 to 103:

[0074] Step 101, obtaining parameter values of a plurality of parameter types of a plurality of battery cells of an energy storage system; the plurality of parameter types include a plurality of types in the following types: battery cell capacity, battery cell voltage, battery cell temperature, and battery cell resistance;

[0075] Step 102, based on the parameter values of the plurality of parameter types, constructing a first input feature map corresponding to each of the plurality of parameter types; wherein the parameter values in the first input feature map are of the same parameter type;

[0076] Step 103, inputting the constructed first input feature map into a pre-trained AI model to predict the initial capacity of the energy storage system and determine the predicted value of the initial capacity.

[0077] It can be understood that in the embodiment of the present application, a method for predicting the initial capacity of an energy storage system before a plurality of battery cells are integrated into the energy storage system is provided, which fills the technical gap of predicting the initial capacity of the energy storage system. The method predicts the predicted value of the initial capacity of the energy storage system based on the first input feature map of a plurality of parameter values of the same parameter type of a plurality of battery cells. In this way, without integrating the plurality of battery cells into the energy storage system, the initial capacity of the energy storage system is measured and obtained, thereby benefiting the manufacturer of the energy storage system to determine the number of battery cells of the to-be-integrated energy storage system for reference based on the predicted value with high accuracy, and further benefiting to reduce the hardware cost of the energy storage system while avoiding the problem of customer complaints due to the initial capacity not meeting the customer's requirement.

[0078] It can be understood that in the embodiments of the present application, the inventors of the present application find that when a plurality of battery cells are integrated into an energy storage system, the battery cell capacity, battery cell voltage, battery cell temperature or battery cell resistance of the plurality of battery cells will interact with each other, thereby affecting the initial capacity of the energy storage system. Therefore, in the embodiments of the present application, based on the first input feature map including the above-mentioned parameter types, the initial capacity of the energy storage system is determined, which is beneficial to improve the accuracy of the predicted value of the predicted initial capacity by determining the key factors affecting the initial capacity of the energy storage system.

[0079] The further optional embodiments of the above respective steps and related terms are described below.

[0080] In step 101, parameter values of a plurality of parameter types of a plurality of battery cells of an energy storage system are obtained; the plurality of parameter types include one or more of the following types: battery cell capacity, battery cell voltage, battery cell temperature, battery cell resistance.

[0081] It should be understood that in the embodiments of the present application, the energy storage system is not limited, and the energy storage system refers to a device for storing and releasing electrical energy, and the energy storage system is integrated by a plurality of battery cells. In some embodiments, the energy storage system is a battery box, a battery cabinet, etc.

[0082] It should be understood that in the embodiments of the present application, the parameter types of the battery cell are not limited; wherein the battery cell parameters are key indicators for measuring the performance and characteristics of the battery cell. In some embodiments, the parameter types further include but are not limited to at least one of the following: battery cell energy, battery cell charge and discharge rate, battery cell cycle life, battery cell self-discharge rate, battery cell operating temperature range, battery cell maximum charging current, battery cell maximum discharging current, battery cell overcharge protection voltage, battery cell overdischarge protection voltage, battery cell thermal runaway temperature, battery cell chemical type, etc.

[0083] In step 102, based on the parameter values of the plurality of parameter types, a first input feature map corresponding to each of the plurality of parameter types is constructed; wherein the parameter values in the first input feature map are of the same parameter type.

[0084] It should be understood that in the embodiments of the present application, the input feature map is not limited, and the input feature map is the initial input data of the convolutional neural network, which is usually a multi-dimensional data, for example: two-dimensional data or three-dimensional array or higher-dimensional tensor.

[0085] In some embodiments, the constructed first input feature map includes one or more of the following first input feature maps: a first input feature map of the battery cell capacity; a first input feature map of the battery cell voltage; a first input feature map of the battery cell temperature; a first input feature map of the battery cell resistance.

[0086] It should be understood that in the embodiments of the present application, the first input feature map of the cell capacity can be understood as including a first number of parameter values of the cell capacity in the first input feature map; the first input feature map of the cell voltage can be understood as including a first number of parameter values of the cell voltage in the first input feature map; the first input feature map of the cell temperature can be understood as including a first number of parameter values of the cell temperature in the first input feature map; and the first input feature map of the cell resistance can be understood as including a first number of parameter values of the cell resistance in the first input feature map.

[0087] It can be understood that in the embodiments of the present application, for the constructed first input feature map, the following two combination schemes are included but not limited to:

[0088] Combination scheme 1: the constructed first input feature map includes: the first input feature map of the cell capacity; the first input feature map of the cell voltage; and the first input feature map of the cell temperature.

[0089] Combination scheme 2: the constructed first input feature map includes: the first input feature map of the cell capacity; the first input feature map of the cell voltage; the first input feature map of the cell temperature; and the first input feature map of the cell resistance.

[0090] It can be understood that in the embodiments of the present application, the more the number of the first input feature maps, the more the types of parameters participating in the prediction of the initial capacity of the energy storage system. Then, the learning of the corresponding relationship between the parameters of the plurality of cells and the initial capacity is more accurate when predicting the initial capacity of the energy storage system based on the first input feature map, so that the predicted value of the initial capacity of the energy storage system is more accurate.

[0091] In the embodiments of the present application, the number of cells of the energy storage system is not limited, and the number of cells is greater than 1. In the embodiments of the present application, the arrangement of the parameter values of the same parameter type of the cells of the energy storage system in the first input feature map is not limited.

[0092] In some embodiments, the energy storage system includes a first number of cells; the first input feature map includes parameter values of the same parameter type of the first number of cells; and the parameter values in the first input feature map are arranged in order of the cell capacity of the first number of cells.

[0093] It can be understood that, in the embodiment of the present application, the parameter values in the first input feature map are arranged in order according to the sizes of the battery capacities of the first number of battery cells, because the inventors of the present application have found through experiments that arranging the parameter values in the first input feature map according to the sizes of the battery capacities of the battery cells can improve the accuracy of the predicted value of the determined initial capacity of the energy storage system, thereby benefiting the manufacturer of the energy storage system to determine the number of battery cells required for the to-be-integrated energy storage system based on the predicted value with high accuracy, and further benefiting to ensure that the initial capacity of the energy storage system integrated based on multiple battery cells meets the customer's requirements while reducing the hardware cost of the energy storage system.

[0094] It should be understood that, in the embodiment of the present application, when the parameter values in the first input feature map are arranged according to the sizes of the battery capacities of the first number of battery cells, the order of arrangement is not limited. In some embodiments, the parameter values in the first input feature map are arranged in order from small to large from left to right and / or from top to bottom according to the sizes of the battery capacities of the first number of battery cells. In other embodiments, the parameter values in the first input feature map are arranged in order from large to small from left to right and / or from top to bottom according to the sizes of the battery capacities of the first number of battery cells.

[0095] In some embodiments, the energy storage system includes a first number of battery cells; the first input feature map includes parameter values of the same parameter type of the first number of battery cells; and the parameter values in the first input feature map are arranged according to the planned arrangement positions of the first number of battery cells.

[0096] In the embodiment of the present application, the planned arrangement position can be understood as the actual arrangement position of the first number of battery cells when integrated into an energy storage system. It should be understood that, in the embodiment of the present application, the planned arrangement position of the first number of battery cells is not limited, i.e., the actual arrangement position of the first number of battery cells is not limited, and the actual arrangement position of the first number of battery cells can be arranged arbitrarily.

[0097] In the embodiment of the present application, the arrangement manner of the parameter values in the first input feature map does not affect the arrangement manner of the first number of battery cells when the first number of battery cells are integrated into an energy storage system. That is, even if the parameter values in the first input feature map are arranged in order according to the sizes of the battery capacities of the first number of battery cells, the first number of battery cells can be arranged arbitrarily when the first number of battery cells are integrated into an energy storage system.

[0098] Exemplarily, in a possible implementation, the parameters of a single battery cell are of various types, such as battery cell voltage, battery cell temperature, battery cell capacity, battery cell resistance, and the like. The energy storage system is composed of several hundred battery cells, thereby forming a high-dimensional array, each row of the array representing a battery cell of a different battery cell number (unique identification of the battery cell), and each column of the array identifying a dimension of a battery cell parameter, such as battery cell voltage. Figure 2 As shown in FIG. 2, 201 and 202 represent battery cells of different battery cell numbers, and each "slice layer" from the outside to the inside represents a type of battery cell parameter, such as 203 representing battery cell voltage and 204 representing battery cell capacity. All battery cells in the energy storage system are arranged in sequence from left to right and from top to bottom on each "slice layer" (i.e., an example of a first input feature map). Figure 3 A schematic diagram of a first input feature map provided by an embodiment of the present application is shown in FIG. 3. Figure 3 As shown in FIG. 3, 301 represents a parameter value of a battery cell in the first input feature map, and Figure 3 As shown in FIG. 3, the parameter type of each battery cell parameter value is the same.

[0099] In step 103, the constructed first input feature map is input into the pre-trained AI model to predict the initial capacity of the energy storage system and determine the predicted value of the initial capacity.

[0100] In an embodiment of the present application, the pre-trained AI model is not limited, and the pre-trained AI model includes a machine learning model and / or a deep learning model. In some embodiments, the pre-trained machine learning model includes a random forest, a support vector machine, a gradient boosting regression, an extreme gradient boosting, an adaptive boosting, and the like. In some embodiments, the pre-trained deep learning model includes, but is not limited to, at least one of a recurrent convolutional neural network, a generative adversarial convolutional neural network, a long short-term memory network, a feedforward neural network, a multilayer perceptron, a convolutional neural network including an attention mechanism, a deep belief network, and the like. In an embodiment of the present application, the training method of the pre-trained AI model is not limited, and the platform on which the pre-trained AI model is deployed is not limited.

[0101] In some embodiments, the structure of the pre-trained AI model includes a first convolutional layer, a first activation layer, a second convolutional layer, a second activation layer, a first pooling layer, a third convolutional layer, a third activation layer, a fourth convolutional layer, a fourth activation layer, a second pooling layer, and a fully connected layer connected in sequence.

[0102] It can be understood that in the embodiments of the present application, the inventors of the present application have verified through experiments that setting the structure of the AI model as the first convolutional layer, the first activation layer, the second convolutional layer, the second activation layer, the first pooling layer, the third convolutional layer, the third activation layer, the fourth convolutional layer, the fourth activation layer, the second pooling layer and the full connection layer in sequence is beneficial to improving the accuracy of the predicted value of the predicted initial capacity.

[0103] It should be understood that in the embodiments of the present application, the size of the convolution kernel, the number of convolution kernels, the step, the number of padding, the number of channels and the like parameters of the convolution kernel in the first convolutional layer, the second convolutional layer, the third convolutional layer and the fourth convolutional layer are not limited. In some embodiments, the number of convolution kernels in the first convolutional layer, the second convolutional layer and the third convolutional layer is different, the number of channels of the first convolutional layer and the second convolutional layer is the same, and the number of channels of the third convolutional layer and the fourth convolutional layer is the same.

[0104] In the embodiments of the present application, the activation function used by the first activation layer, the second activation layer, the third activation layer and the fourth activation layer is not limited, and the activation function used by the first activation layer, the second activation layer, the third activation layer and the fourth activation layer can be the same or different. In some embodiments, the activation function used by each activation layer includes but is not limited to at least one of the following: S-type function (Sigmoid function), hyperbolic tangent function (Tanh function), linear rectification function (ReLU function), Leaky ReLU function, variant of ReLU function, etc.

[0105] It should be understood that in the embodiments of the present application, the pooling operation used by the first pooling layer and the second pooling layer is not limited, and the pooling operation used by the first pooling layer and the second pooling layer can be the same or different. In some embodiments, the pooling operation used by each pooling layer includes but is not limited to at least one of the following: maximum pooling (Max Pooling), average pooling (Average Pooling), global maximum pooling (Global Max Pooling), global average pooling (Global Average Pooling), L2 pooling (L2 Pooling), overlapping pooling (Overlapping Pooling), stochastic pooling (Stochastic Pooling), mixed pooling (Mixed Pooling), etc.

[0106] Exemplarily, in a possible implementation manner, Figure 4 A schematic diagram of a energy storage system provided by the embodiments of the present application is shown in FIG. 4. Figure 4 As shown in FIG. 4, 401 represents a single cell. Figure 5 A schematic diagram of a mathematical model of a energy storage system provided by the embodiments of the present application is shown in FIG. 5. Figure 5As shown, the X-axis and the Y-axis represent the cell position, and the Z-axis represents the parameter type of the cell (i.e., the cell key parameter in the figure). Based on the mathematical model, a three-dimensional matrix of the energy storage system can be obtained, for example: [[[3.287911, 313361.2, 28.897, 1366.3, 420.31],..., [3.288811, 313470.9, 31.283, 1366.3, 420.31]]..., [[3.287911, 313361.2, 28.897, 1366.3, 420.31],..., [3.288811, 313470.9, 31.283, 1366.3, 420.31]]]. The parameters of the cells are sorted according to the size of the cell capacity of the single cell from left to right and from top to bottom, and the three-dimensional matrix is reconstructed to obtain a three-dimensional matrix of [20, 20, 5] (i.e., an example of the first input feature map).

[0107] Figure 6 A structure diagram of an AI model provided for an embodiment of the present application is shown in FIG. 2. Figure 6 As shown, the constructed three-dimensional matrix [20, 20, 5] is input to a convolution layer with a convolution kernel size of 3x3, a convolution kernel number of 5, a step length of 1, a padding number of 1, and a channel number of 64, and an activation layer with a ReLU activation function, to obtain a feature matrix of [20, 20, 64]; the feature matrix [20, 20, 64] is input to a convolution layer with a convolution kernel size of 3x3, a convolution kernel number of 64, a step length of 1, a padding number of 1, and a channel number of 64, and an activation layer with a ReLU activation function, to obtain a feature matrix of [20, 20, 64]; the feature matrix [20, 20, 64] is input to a pooling layer with a pooling window of 3x3, to perform a maximum pooling operation, to output a feature matrix of [18, 18, 64]; the feature matrix [18, 18, 64] is input to a convolution layer with a convolution kernel size of 3x3, a convolution kernel number of 5, a step length of 1, a padding number of 1, and a channel number of 128, and an activation layer with a ReLU activation function, to obtain a feature matrix of [18, 18, 128]; the feature matrix [18, 18, 128] is input to a convolution layer with a convolution kernel size of 3x3, a convolution kernel number of 64, a step length of 1, a padding number of 1, and a channel number of 128, and an activation layer with a ReLU activation function, to obtain a feature matrix of [18, 18, 128]; the feature matrix [18, 18, 128] is input to a pooling layer with a pooling window of 3x3, to perform a maximum pooling operation, to obtain a feature matrix of [16, 16, 128]; and the feature matrix [16, 16, 128] is input to a fully connected layer, to obtain a prediction value.

[0108] Figure 7 A schematic diagram of a convolution operation provided for an embodiment of the present application is shown in FIG. 3.Figure 7 As shown, 701 represents the input, 702 represents the convolution kernel, and 703 represents the output result. Figure 8 This is a schematic diagram of a pooling operation provided in an embodiment of this application, as shown below. Figure 8 As shown, 801 represents the input, 802 represents the max pooling operation with a window size of 2×2, and 803 represents the output of the max pooling operation. Figure 9 This is a schematic diagram of an activation function provided in an embodiment of this application, such as... Figure 9 As shown, when the input z is less than or equal to 0, the output is 0; when the input z is greater than 0, the output is z; where the horizontal axis represents the input value and the vertical axis represents the output value.

[0109] Figure 10 This is a schematic diagram of the backpropagation process of a convolutional neural network provided in an embodiment of this application, as shown below. Figure 10 As shown, the convolution kernel matrix is ​​the matrix used to extract features in a convolutional neural network. The convolution kernel slides across the input data, generating feature maps through convolution operations. The true value is the target output, and the predicted value is the output generated by the model based on the input data. The difference refers to the discrepancy between the true and predicted values, used to calculate the loss function and then for backpropagation. The learning rate coefficient is a hyperparameter in the optimization algorithm, used to control the step size of weight updates. The gradient is the partial derivative of the loss function with respect to the model parameters. The formula for calculating the gradient is:

[0110]

[0111] in, It is the first l +1 floor Each convolutional kernel is located at... The weight, The loss function E is relative to the first l +1 layer activation value In position ( ) and the The gradient of each convolutional kernel, Is the activation function in The derivative at point, Indicates the size of the convolution kernel. Indicates step length (i.e. stride). Indicates the first l The number of convolutional kernels in a layer.

[0112] The update formula for the convolution kernel is:

[0113]

[0114] in, It is the first l Layer weights is the weight of the l +1 layer, denotes a learning rate, denotes a gradient of the loss function E with respect to the k convolution kernel weight of the +1 layer, l denotes an activation value of the +1 layer, l denotes a gradient of the loss function E with respect to the +1 layer activation value.

[0115] In some embodiments, the learning rate coefficient is set to 0.1.

[0116] In some embodiments, the first input feature map constructed is input into a pre-trained AI model to predict the initial capacity of the energy storage system, and a predicted value of the initial capacity is determined. Before the energy storage system is integrated, the first input feature map constructed is input into the pre-trained AI model to predict the initial capacity of the energy storage system, and the predicted value of the initial capacity is determined.

[0117] It can be understood that in the embodiments of the present application, the initial capacity of the energy storage system can be predicted based on the plurality of first input feature maps before the plurality of battery cells are integrated into the energy storage system, so as to determine the predicted value of the initial capacity. In this way, the initial capacity of the energy storage system does not need to be obtained after the plurality of battery cells are integrated into the energy storage system, thereby benefiting the manufacturer of the energy storage system to determine the number of battery cells of the energy storage system to be integrated based on the predicted value with high accuracy, and thereby benefiting to reduce the hardware cost of the energy storage system while avoiding complaints due to the initial capacity not meeting the requirements of customers.

[0118] In some embodiments, the energy storage system initial capacity prediction method further includes: recommending a target number of battery cells required for integrating the energy storage system based on a size relationship between the predicted value of the initial capacity and a preset expected value.

[0119] It can be understood that in the embodiments of the present application, the target number of battery cells required for integrating the energy storage system is recommended based on the size relationship between the predicted value of the initial capacity and the preset expected value. In this way, the target number of battery cells required for integrating the energy storage system can be automatically or intelligently recommended, thereby improving the efficiency of determining the target number of battery cells required for integrating the energy storage system and reducing the labor cost of determining the target number of battery cells required for integrating the energy storage system.

[0120] In some embodiments, the target number of battery cells required by the integrated energy storage system is recommended based on a size relationship between the predicted value of the initial capacity and the preset expected value, including: in the case that the predicted value of the initial capacity meets a first condition, recommending the target number as a first number; wherein the first number is the number of battery cells in the energy storage system; the first condition includes that the predicted value of the initial capacity is greater than the preset expected value, and the difference between the predicted value of the initial capacity and the preset expected value is within a first numerical range.

[0121] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity is greater than the preset expected value and the difference between the predicted value of the initial capacity and the preset expected value is within the first numerical range, the target number is recommended as the first number. That is, when the part of the predicted value of the initial capacity that is higher than the preset expected value is within the acceptable overrun range, the target number is recommended as the first number, so that the recommended target number is more reasonable, and then when the energy storage system is integrated based on the reasonable number of battery cells, the initial capacity of the energy storage system can meet the requirements of the customer while reducing the hardware cost of the energy storage system.

[0122] It should be understood that in the embodiments of the present application, the first numerical range is not limited. In some embodiments, the first numerical range is preset. In other embodiments, the first numerical range can be determined according to various factors such as customer preferences, application scenarios of the energy storage system, prices of the energy storage system, prediction accuracy of the pre-trained AI model, or prediction error of the pre-trained AI model. In one possible implementation, the first numerical range is [5, 10].

[0123] In some embodiments, Figure 11 An implementation process schematic diagram for recommending a target number of battery cells required by an integrated energy storage system is provided for the embodiments of the present application, as shown in Figure 11 The target number of battery cells required by the integrated energy storage system can be recommended by the following steps 1101 and 1102:

[0124] Step 1101, in the case that the predicted value of the initial capacity does not meet the first condition, determining a plurality of second input feature maps; wherein the second input feature map includes parameter values of a second number of battery cells, and the second number is different from the first number; the parameter values in the second input feature map are of the same parameter type; the same element position of different second input feature maps corresponds to different parameter types of the same battery cell;

[0125] Step 1102, based on the plurality of second input feature maps and the pre-trained AI model, re-determining the predicted value of the initial capacity until the predicted value of the initial capacity meets the first condition, and recommending the target number as the number of battery cells corresponding to the predicted value that meets the first condition.

[0126] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity does not satisfy the first condition, a plurality of second input feature maps of the parameter value of the second number of battery cells are determined, and the second number is different from the first number. That is, in the case that the predicted value of the initial capacity does not satisfy the first condition, the number of battery cells of the integrated energy storage system is updated, the predicted value of the initial capacity is re-predicted until the predicted value of the initial capacity satisfies the first condition, and then the target number is recommended as the number of battery cells corresponding to the predicted value satisfying the first condition. That is, when it is ensured that the part of the predicted value of the initial capacity being higher than the preset expected value is within the acceptable overrun range, the target number of battery cells required by the integrated energy storage system is recommended. In this way, the recommended target number is more reasonable, that is, the initial capacity of the energy storage system integrated based on the target number of battery cells can meet the customer's requirements while reducing the hardware cost of the energy storage system.

[0127] It should be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity does not satisfy the first condition, the manner of determining the number of battery cells of the integrated energy storage system is not limited. In some embodiments, in the case that the predicted value of the initial capacity does not satisfy the first condition, a change amount of the number of battery cells of the integrated energy storage system is determined; and based on the change amount, the number of battery cells of the integrated energy storage system is determined.

[0128] In some embodiments, in the case that the predicted value of the initial capacity does not satisfy the first condition, determining a change amount of the number of battery cells of the integrated energy storage system comprises: in the case that the predicted value of the initial capacity does not satisfy the first condition, determining that the change amount of the number of battery cells of the integrated energy storage system is a preset change amount.

[0129] It should be understood that in the embodiments of the present application, the preset change amount is not limited. In some embodiments, the preset change amount can be determined according to an empirical value. In other embodiments, the preset change amount can be determined according to the difference between the predicted value of the initial capacity and the preset expected value, the relationship between the difference and the first numerical range, and the capacity of a single battery cell. In one possible implementation, the preset change amount is 1.

[0130] In other embodiments, in the case that the predicted value of the initial capacity does not satisfy the first condition, a change amount of the number of battery cells of the integrated energy storage system is determined based on the difference between the predicted value of the initial capacity and the preset expected value, the relationship between the difference and the first numerical range, and the capacity of a single battery cell.

[0131] Exemplarily, in one possible implementation, the predicted value of the initial capacity is 98, the preset expected value is 100, the capacity value of a single battery cell is 1, the first numerical range is [5, 10], and the change amount of the number of battery cells of the integrated energy storage system is determined as any integer in 7-12.

[0132] It should be understood that in the embodiments of the present application, the "until the predicted value of the initial capacity meets the first condition, the recommended target number is the number of battery cells corresponding to the predicted value meeting the first condition" can be understood as: in the case that the predicted value of the re-determined initial capacity meets the first condition, the recommended target number is the second number; in the case that the re-determined initial capacity does not meet the first condition, the second input feature map is taken as the first input feature map, and the steps of determining the plurality of second input feature maps in the case that the predicted value of the initial capacity does not meet the first condition, and re-determining the predicted value of the initial capacity based on the plurality of second input feature maps and the pre-trained AI model are performed.

[0133] Exemplarily, in a possible implementation, the first input feature map includes parameter values of 100 battery cells, the predicted value of the initial capacity is 98, the preset expected value is 100, the capacity value of a single battery cell is 1, the first numerical range is [5, 10], and the determined second input feature map includes parameter values of 107 (i.e., an example of the second number) battery cells. Based on the plurality of second input feature maps and the pre-trained AI model, the re-determined predicted value of the initial capacity is 104. 104 does not meet the first condition, the second input feature map is taken as the first input feature map, the second input feature map is re-determined, and the re-determined second input feature map includes parameter values of 110 battery cells. Based on the plurality of second input feature maps and the pre-trained AI model, the re-determined predicted value of the initial capacity is 107, which meets the first condition, and the recommended target number is 107.

[0134] In some embodiments, the predicted value of the initial capacity not meeting the first condition includes that the predicted value of the initial capacity is less than or equal to the preset expected value, and the second number is greater than the first number.

[0135] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity does not meet the preset expected value or just meets the preset expected value, the number of battery cells of the to-be-integrated energy storage system is increased, and then the predicted value of the initial capacity is re-determined based on the plurality of second input feature maps with the increased number of battery cells and the pre-trained AI model. In this way, it is beneficial to avoid the problem that the target number of battery cells integrated in the recommended energy storage system does not meet the requirements of the customer, thereby causing customer complaints.

[0136] Exemplarily, in a possible implementation, the predicted value of the initial capacity is 98, and the preset expected value is 100, and the number of battery cells of the to-be-integrated energy storage system is increased.

[0137] In some embodiments, the first condition not being met by the predicted value of the initial capacity includes that the predicted value of the initial capacity is greater than the preset expected value, and a difference between the predicted value of the initial capacity and the preset expected value is less than a lower limit value of the first numerical range; and the second quantity is greater than the first quantity.

[0138] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity is higher than the preset expected value, but the part that is higher is not too much, the number of the battery cells of the to-be-integrated energy storage system is increased, and then the predicted value of the initial capacity is re-determined based on the plurality of second input feature maps with the increased number of battery cells and the pre-trained AI model. In this way, it is beneficial to ensure that the energy storage system integrated with the target number of battery cells meets the requirements of the customer, thereby avoiding the problem of customer complaints.

[0139] Exemplarily, in a possible implementation, in the case that the predicted value of the initial capacity is 102 and the preset expected value is 100, the number of the battery cells of the to-be-integrated energy storage system is increased.

[0140] In some embodiments, the first condition not being met by the predicted value of the initial capacity includes that the predicted value of the initial capacity is greater than the preset expected value, and a difference between the predicted value of the initial capacity and the preset expected value is greater than an upper limit value of the first numerical range; and the second quantity is less than the first quantity.

[0141] It can be understood that in the embodiments of the present application, in the case that the predicted value of the initial capacity far exceeds the preset expected value, the number of the battery cells of the to-be-integrated energy storage system is reduced, and then the predicted value of the initial capacity is re-determined based on the plurality of second input feature maps with the reduced number of battery cells and the pre-trained AI model. In this way, it is beneficial to avoid that the energy storage system integrated with the target number of battery cells is far higher than the requirements of the customer, thereby reducing the hardware cost of the energy storage system.

[0142] Exemplarily, in a possible implementation, in the case that the predicted value of the initial capacity is 118 and the preset expected value is 100, the number of the battery cells of the to-be-integrated energy storage system is reduced.

[0143] Based on this, the exemplary application of the embodiments of the present application in an actual application scenario will be described below.

[0144] The embodiments of the present application provide a method for predicting an initial capacity of an energy storage system, which can screen key factors affecting the initial capacity of an electric cabinet or an electric box.

[0145] The parameters related to each battery cell in a group of battery cells, i.e., the parameters affecting the initial capacity, are extracted by using a convolutional neural network framework, so as to solve the selection problem of the key factors affecting the initial capacity of the energy storage system after the integration of multiple battery cells.

[0146] The calculation of the convolutional layer has two key steps, one is local correlation, and the other is window sliding. The local correlation is to regard each neuron as a filter, and the window sliding is to calculate the local data by the filter. The implementation of convolution involves three parameters: stride (an example of step length), filter size, and padding number. The calculation process of convolution is that the stride controls the filter to perform convolution operation, for example: the size of the input image is 7x7, the filter size is 3x3, and the stride is 1. The change of the stride causes the change of the receptive field. In biology, the receptive field usually refers to the area in which a sensory neuron (such as a retinal neuron or a visual neuron in the cerebral cortex) reacts to light stimuli in a certain part of the visual field. In the computer field, the receptive field is used to describe which area of the input image a neuron in a certain layer of the network is sensitive to. The size of the receptive field determines how much context information the neuron can capture. The larger the convolution kernel (for example: 3x3), the larger the receptive field and the more context information captured. Convolution calculation reduces the spatial dimension of data. In the early layers of the network, if it is desired to retain more information of the original input content to facilitate the extraction of the features of the lower layer, padding can be used, for example: for an input image of 32x32x3, a filter of 5x5x5 is used for convolution, and the output is 28x28x3. In order to make the output image dimension 32x32x3, the input image can be padded with 0 outside the input image. After applying zero padding of size 2, the input image is updated to 36x36x3. Thus, using a filter of 5x5x3 can still obtain an output of 32x32x3.

[0147] The role of the activation layer is to make a nonlinear mapping of the convolution result; the pooling layer is sandwiched between consecutive convolution layers, used to compress the amount of data and parameters, and reduce overfitting. In short, if the input is an image, the main role of the pooling layer is to compress the image.

[0148] The specific role of the pooling layer includes feature invariance and feature dimension reduction. The feature invariance can be understood as that when the image is compressed, only some unimportant information is removed, and important features are retained, for example, after a picture of a small dog is reduced, we can still recognize that the image is a dog. Feature dimension reduction is to remove redundant features and extract important features. To some extent, it prevents overfitting and facilitates optimization.

[0149] Full connection layer refers to that all neurons between two layers have weight connection, and the full connection layer is usually at the tail of the convolutional neural network. That is, the connection mode of the neurons is the same as that of the traditional neural network.

[0150] It can be understood that in the embodiments of the present application, the extraction of the key factors affecting the initial capacity of the electric cabinet or the electric box is realized, and on the basis of lacking mechanism guidance, the single battery key parameters and the multi-battery coupling information are fused for sufficient feature extraction. And through the gradient descent back calculation to continuously optimize the parameters of the neural network, and then realize the supervised learning, realize the prediction of the initial capacity of the electric cabinet or the electric box.

[0151] In a possible implementation, the parameters of a single battery have multiple types, such as battery voltage, battery temperature, battery capacity, battery resistance, etc. The energy storage system is composed of hundreds of batteries, thereby forming a high-dimensional array, each row of the array represents a battery of different battery number (unique identification of the battery), and each column of the array identifies a certain dimension (i.e. a certain type) of the battery parameter, such as battery voltage. The mathematical model of the energy storage system can be abstracted as a “cube” (a three-dimensional data matrix), as shown in Figure 2 201 and 202 represent batteries of different battery numbers, and each “slice layer” from the outside to the inside represents a type of battery parameter, such as 203 representing battery voltage and 204 representing battery capacity. All batteries in the energy storage system are arranged in order from left to right and from top to bottom on each “slice layer”.

[0152] It can be understood that in the embodiments of the present application, the supervised machine learning feature extraction work in the environment of mutual coupling of a complex system (i.e. an example of an energy storage system) is realized, and the bottleneck that the related machine learning model and the deep learning model cannot find effective feature factors is solved. The method has relatively effective prediction effect in the complex interactive environment.

[0153] In a possible implementation, Figure 4 A schematic diagram of an energy storage system provided by the embodiments of the present application is shown in Figure 4 401 represents a single battery. Figure 5 A schematic diagram of a mathematical model of an energy storage system provided by the embodiments of the present application is shown in Figure 5As shown, the X-axis and the Y-axis represent the cell position, and the Z-axis represents the parameter type of the cell (i.e., the cell key parameter in the figure). Based on the mathematical model, a three-dimensional matrix of the energy storage system can be obtained, for example: [[[3.287911, 313361.2, 28.897, 1366.3, 420.31],... [3.288811, 313470.9, 31.283, 1366.3, 420.31]]... [[3.287911, 313361.2, 28.897, 1366.3, 420.31],... [3.288811, 313470.9, 31.283, 1366.3, 420.31]]]. The parameters of the cells are sorted according to the size of the cell capacity of the single cell from left to right and from top to bottom, and the three-dimensional matrix is reconstructed to obtain a three-dimensional matrix of [20, 20, 5] (i.e., an example of the first input feature map).

[0154] Figure 6 A structure schematic diagram of an AI model provided for an embodiment of the present application is as shown in Figure 6 The constructed three-dimensional matrix [20, 20, 5] is input to a convolution layer with a convolution kernel size of 3x3, a convolution kernel number of 5, a step length of 1, a padding number of 1, a channel number of 64, and an activation function of ReLU, to obtain a feature matrix of [20, 20, 64]; the feature matrix [20, 20, 64] is input to a convolution layer with a convolution kernel size of 3x3, a convolution kernel number of 64, a step length of 1, a padding number of 1, a channel number of 64, and an activation function of ReLU, to obtain a feature matrix of [20, 20, 64]; the feature matrix [20, 20, 64] is input to a pooling layer with a pooling window of 3x3, to perform a maximum pooling operation, to output a feature matrix of [18, 18, 64]; the feature matrix [18, 18, 64] is input to a convolution layer with a convolution kernel size of 3x3, a convolution kernel number of 128, a step length of 1, a padding number of 1, a channel number of 128, and an activation function of ReLU, to obtain a feature matrix of [18, 18, 128]; the feature matrix [18, 18, 128] is input to a convolution layer with a convolution kernel size of 3x3, a convolution kernel number of 128, a step length of 1, a padding number of 1, a channel number of 128, and an activation function of ReLU, to obtain a feature matrix of [18, 18, 128]; the feature matrix [18, 18, 128] is input to a pooling layer with a pooling window of 3x3, to perform a maximum pooling operation, to obtain a feature matrix of [16, 16, 128]; and the feature matrix [16, 16, 128] is input to a fully connected layer, to obtain a prediction value.

[0155] Figure 10This is a schematic diagram of the backpropagation process of a convolutional neural network provided in an embodiment of this application, as shown below. Figure 10 As shown, the convolution kernel matrix is ​​the matrix used to extract features in a convolutional neural network. The convolution kernel slides across the input data, generating feature maps through convolution operations. The true value is the target output, and the predicted value is the output generated by the model based on the input data. The difference refers to the discrepancy between the true and predicted values, used to calculate the loss function and then for backpropagation. The learning rate coefficient is a hyperparameter in the optimization algorithm, used to control the step size of weight updates. The gradient is the partial derivative of the loss function with respect to the model parameters.

[0156] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of this application without departing from the spirit and scope of the technical solutions of this application, and all such modifications or substitutions should be covered within the protection scope of this application.

[0157] It should be noted that although the steps of the method in this application are described in a specific order in the accompanying drawings, this does not require or imply that these steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps; or steps from different embodiments may be combined into a new technical solution. Based on the foregoing embodiments, this application provides an apparatus comprising the included modules and the units included in each module, which can be implemented by a processor; of course, it can also be implemented by specific logic circuits; in implementation, the processor can be an AI acceleration engine (such as an NPU), a graphics processing unit (GPU), a central processing unit (CPU), a microprocessor (MPU), a digital signal processor (DSP), or a field-programmable gate array (FPGA), etc.

[0158] Figure 12 This is a schematic diagram of the structure of an initial capacity prediction device for an energy storage system provided in an embodiment of this application, as shown below. Figure 12 As shown, the initial capacity prediction device 120 for an energy storage system includes an acquisition module 1201, a construction module 1202, and a prediction module 1203, wherein:

[0159] The acquisition module 1201 is configured to acquire parameter values ​​of multiple parameter types from multiple cells in the energy storage system; the multiple parameter types include several of the following types: cell capacity, cell voltage, cell temperature, and cell resistance;

[0160] The constructing module 1202 is configured to construct a first input feature map corresponding to each of the plurality of parameter types based on the parameter values of the plurality of parameter types; wherein the parameter values in the first input feature map are of the same parameter type.

[0161] The predicting module 1203 is configured to input the constructed first input feature map into the pre-trained AI model to predict the initial capacity of the energy storage system and determine a predicted value of the initial capacity.

[0162] In some embodiments, the energy storage system includes a first number of battery cells; the first input feature map includes parameter values of the same parameter type of the first number of battery cells; and the parameter values in the first input feature map are arranged in order of the cell capacities of the first number of battery cells.

[0163] In some embodiments, the predicting module 1203 is configured to input the constructed first input feature map into the pre-trained AI model to predict the initial capacity of the energy storage system and determine a predicted value of the initial capacity before the energy storage system is integrated.

[0164] In some embodiments, the energy storage system initial capacity prediction apparatus 120 further includes a recommending module; wherein the recommending module is configured to recommend a target number of battery cells required for the integrated energy storage system based on a size relationship between the predicted value of the initial capacity and a preset expected value.

[0165] In some embodiments, the recommending module is configured to recommend the target number as a first number in a case where the predicted value of the initial capacity satisfies a first condition; wherein the first number is the number of battery cells in the energy storage system; and the first condition includes that the predicted value of the initial capacity is greater than the preset expected value and a difference between the predicted value of the initial capacity and the preset expected value is within a first numerical range.

[0166] In some embodiments, the recommending module is configured to determine a plurality of second input feature maps in a case where the predicted value of the initial capacity does not satisfy the first condition; wherein the second input feature maps include parameter values of a second number of battery cells, the second number being different from the first number; the parameter values in the second input feature maps are of the same parameter type; a same element position of different second input feature maps corresponds to different parameter types of a same battery cell; and the predicted value of the initial capacity is re-determined based on the plurality of second input feature maps and the pre-trained AI model until the predicted value of the initial capacity satisfies the first condition, and the target number is recommended as the number of battery cells corresponding to the predicted value satisfying the first condition.

[0167] In some embodiments, the predicted value of the initial capacity not satisfying the first condition includes that the predicted value of the initial capacity is less than or equal to the preset expected value; and the second number is greater than the first number.

[0168] In some embodiments, the first condition not being satisfied by the predicted value of the initial capacity comprises: the predicted value of the initial capacity being greater than the preset expected value, and a difference between the predicted value of the initial capacity and the preset expected value being less than a lower limit value of the first numerical range; and the second quantity being greater than the first quantity.

[0169] In some embodiments, the first condition not being satisfied by the predicted value of the initial capacity comprises: the predicted value of the initial capacity being greater than the preset expected value, and a difference between the predicted value of the initial capacity and the preset expected value being greater than an upper limit value of the first numerical range; and the second quantity being less than the first quantity.

[0170] The above device embodiments are similar to the descriptions of the above method embodiments, and have similar beneficial effects to the method embodiments. For technical details not disclosed in the device embodiments of the present application, please refer to the description of the method embodiments of the present application.

[0171] It should be noted that the division of modules in the embodiments of the present application is illustrative, and is only a logical functional division. In actual implementation, another division mode can be used. In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or can be physically separated, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware, or in the form of a software functional unit. It can also be realized in the form of a combination of software and hardware.

[0172] It should be noted that, in the embodiments of the present application, if the above method is realized in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing an electronic device to execute all or part of the method of the embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk, and various storage media that can store program codes. Thus, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0173] The embodiments of the present application provide an electronic device, Figure 13 A structural schematic diagram of an electronic device provided by the embodiments of the present application is shown in FIG. 13. Figure 13 As shown in FIG. 13, the electronic device 130 includes a memory 1301 and a processor 1302. The memory 1301 stores a computer program executable on the processor 1302. When the processor 1302 executes the program, the steps in the method provided in the above embodiments are implemented.

[0174] It should be noted that the memory 1301 is configured to store instructions and applications executable by the processor 1302, and can also buffer data to be processed or having been processed by the processor 1302 and each module in the electronic device 130, and can be implemented by a flash memory (FLASH) or a random access memory (RAM).

[0175] In the embodiments of the present application, the type of the electronic device is not limited, and the electronic device can be a smart phone, a notebook computer, a tablet computer, a smart bracelet, an Internet of Things (IoT) device, or a vehicle-mounted device, etc.

[0176] The embodiments of the present application provide a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps in the method provided in the above embodiments.

[0177] The embodiments of the present application provide a computer program product containing instructions, which, when running on a computer, causes the computer to perform the steps in the method provided in the above method embodiments.

[0178] It should be noted that the descriptions of the above storage medium and device embodiments are similar to the descriptions of the above method embodiments, and have similar beneficial effects to the method embodiments. For technical details not disclosed in the storage medium and device embodiments of the present application, please refer to the description of the method embodiments of the present application for understanding.

[0179] It should be understood that the "one embodiment" or "an embodiment" or "some embodiments" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiments are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" or "in some embodiments" appearing throughout the specification does not necessarily mean the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present application, the size of the serial number of the above processes does not mean the execution order, and the execution order of the processes should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The serial number of the above embodiments of the present application is only for description, not representing the advantages and disadvantages of the embodiments. The above description of each embodiment tends to emphasize the differences between each embodiment, and the same or similar parts can be referred to each other. For the sake of brevity, this paper will not repeat here.

[0180] The term "and / or" in this paper is only to describe the association relationship of the associated objects, which means that there can be three kinds of relationships, for example, object A and / or object B, which can represent three cases of existence of object A alone, existence of object A and object B, and existence of object B alone.

[0181] It should be noted that, in the present document, the terms "comprising", "containing" or any other similar term are intended to encompass non-exclusive inclusion, such that processes, methods, articles, or apparatuses that comprise a list of elements are not limited to those elements, but can also include other elements not expressly listed, or inherent to such processes, methods, articles, or apparatuses. Without further limitation, an element defined by the phrase "comprising a" does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0182] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The above-described embodiments are merely illustrative, for example, the division of modules is only a logical function division, and actual implementation can have another division manner, such as: a plurality of modules or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed components can be indirect coupling or communication connection between some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0183] The above-described modules described as separate components can or can not be physically separated, and the components displayed as modules can or can not be physical modules; they can be located in one place or distributed on multiple network units; and some or all of the modules can be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0184] In addition, each functional module in each embodiment of the present application can be integrated in one processing unit, or each module can be a separate unit, or two or more modules can be integrated in one unit; the above integrated modules can be realized in the form of hardware or hardware plus software functional units.

[0185] Those of ordinary skill in the art can understand that all or part of the steps of the above method embodiments can be completed by program instruction related hardware, and the above program can be stored in a computer readable storage medium, and the program executes the steps of the above method embodiments when executed; and the above storage medium includes mobile storage devices, read only memory (Read Only Memory, ROM), magnetic discs or optical discs, and various storage medium that can store program codes.

[0186] Alternatively, the above-mentioned integrated units of the present application, if realized in the form of software function modules and sold or used as independent products, can also be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a plurality of instructions for causing a display device or a cloud device to execute all or part of the embodiments of the method of the present application. The aforementioned storage medium includes mobile storage devices, ROM, magnetic disks or optical disks, and various media that can store program codes.

[0187] The methods disclosed in the several method embodiments of the present application can be combined arbitrarily without conflict to obtain new method embodiments.

[0188] The features disclosed in the several product embodiments of the present application can be combined arbitrarily without conflict to obtain new product embodiments.

[0189] The features disclosed in the several method or device embodiments of the present application can be combined arbitrarily without conflict to obtain new method or device embodiments.

[0190] The above is only an implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for predicting initial capacity of an energy storage system, characterized by, The method comprises: obtaining parameter values of multiple parameter types of multiple battery cells of the energy storage system; the multiple parameter types include multiple types from the following types: battery cell capacity, battery cell voltage, battery cell temperature, battery cell resistance; based on the parameter values of the multiple parameter types, constructing a first input feature map corresponding to each of the multiple parameter types; wherein the parameter values in the first input feature map are of the same parameter type; the first input feature map is two-dimensional data; before the energy storage system integration, inputting the constructed first input feature map into a pre-trained AI model to predict an initial capacity of the energy storage system and determine a predicted value of the initial capacity; based on the size relationship between the predicted value of the initial capacity and a preset expected value, recommending a target number of battery cells required for integrating the energy storage system.

2. The method of claim 1, wherein, The energy storage system includes a first number of battery cells; the first input feature map includes parameter values of the same parameter type of the first number of battery cells; the parameter values in the first input feature map are arranged in order of the size of the battery cell capacity of the first number of battery cells.

3. The method of claim 1, wherein, Based on the size relationship between the predicted value of the initial capacity and a preset expected value, recommending a target number of battery cells required for integrating the energy storage system, comprising: in the case that the predicted value of the initial capacity meets a first condition, recommending the target number as a first number; wherein the first number is the number of battery cells in the energy storage system; the first condition includes that the predicted value of the initial capacity is greater than the preset expected value, and the difference between the predicted value of the initial capacity and the preset expected value is within a first numerical range.

4. The method of claim 3, wherein, Based on the size relationship between the predicted value of the initial capacity and a preset expected value, recommending a target number of battery cells required for integrating the energy storage system, comprising: in the case that the predicted value of the initial capacity does not meet the first condition, determining multiple second input feature maps; wherein the second input feature map includes parameter values of a second number of battery cells, the second number is different from the first number; the parameter values in the second input feature map are of the same parameter type; the same element position of different second input feature maps corresponds to the same battery cell of different parameter types; based on the multiple second input feature maps and the pre-trained AI model, re-determining the predicted value of the initial capacity until the predicted value of the initial capacity meets the first condition, and recommending the target number as the number of battery cells corresponding to the predicted value that meets the first condition.

5. The method of claim 4, wherein, The predicted value of the initial capacity does not meet the first condition includes that the predicted value of the initial capacity is less than or equal to the preset expected value; the second number is greater than the first number.

6. The method of claim 4, wherein, The predicted value of the initial capacity does not meet the first condition includes that the predicted value of the initial capacity is greater than the preset expected value, and the difference between the predicted value of the initial capacity and the preset expected value is less than the lower limit value of the first numerical range; the second number is greater than the first number.

7. The method of claim 4, wherein, The prediction of the initial capacity not satisfying the first condition includes that the prediction of the initial capacity is greater than the preset expected value, and a difference between the prediction of the initial capacity and the preset expected value is greater than an upper limit value of the first numerical range; and the second quantity is less than the first quantity.

8. An energy storage system initial capacity prediction device characterized by comprising: The method comprises: The acquisition module is configured to acquire parameter values of a plurality of parameter types of a plurality of battery cells of the energy storage system; the plurality of parameter types include a plurality of types from the following types: battery cell capacity, battery cell voltage, battery cell temperature, and battery cell resistance; The construction module is configured to construct a first input feature map corresponding to each of the plurality of parameter types based on the parameter values of the plurality of parameter types; the parameter values in the first input feature map are of the same parameter type; and the first input feature map is two-dimensional data. The prediction module is configured to input the constructed first input feature map into a pre-trained AI model to predict an initial capacity of the energy storage system and determine a prediction of the initial capacity before the energy storage system is integrated. The recommendation module is configured to recommend a target quantity of battery cells required for integrating the energy storage system based on a size relationship between the prediction of the initial capacity and a preset expected value.

9. An electronic device comprising a memory and a processor, the memory storing a computer program operable on the processor, characterized in that, The processor executes the program to implement the method in any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method in any one of claims 1 to 7.

11. A computer program product comprising computer programs or instructions, characterized in that, The computer program or instructions are executed by the processor to implement the method in any one of claims 1 to 7.

Images