A battery SOC prediction method, device, monitoring system and storage medium

Abstract

A battery SOC prediction method, device, monitoring system and storage medium, the method comprising: determining the charging and discharging stage of the battery according to the current cumulative charging and discharging duration of the battery; selecting a corresponding pre-trained SOC prediction model according to the charging and discharging stage of the battery; wherein each charging and discharging stage corresponds to an SOC prediction model; inputting the current operating parameters of the battery into the corresponding SOC prediction model to obtain the SOC prediction value of the battery. The scheme fully considers the influence of the cumulative charging and discharging duration on the battery SOC prediction, and different SOC prediction models are pre-trained for different charging and discharging stages of the battery, which can greatly improve the accuracy of the SOC value prediction.

Description

Technical Field

[0001] This application relates to, but is not limited to, the field of state of charge assessment for battery energy storage systems, and particularly to a battery SOC prediction method, apparatus, monitoring system, and storage medium. Background Technology

[0002] With the rapid development of new clean energy sources in recent years, energy storage technology has also seen rapid advancements in the industry. Vanadium redox flow batteries, relying on their excellent stability and safety, have been widely used in the energy storage sector, becoming a mainstream technology solution. Flow batteries utilize the redox reactions generated by the flow of the electrolyte to continuously change the valence state of vanadium ions, thereby achieving the purpose of charging and discharging. During the use of vanadium redox flow batteries, it is necessary to monitor the battery's state of charge (SOC) in real time to avoid overcharging or over-discharging, which could damage the battery. Therefore, accurate real-time prediction of the SOC of vanadium redox flow batteries is crucial for the operation of the system.

[0003] Currently, the mainstream ampere-hour integration method and reference battery method have certain limitations in predicting SOC: the ampere-hour integration method calculates the remaining capacity by integrating the current, which only considers the impact of current as a single factor on SOC, resulting in a large prediction deviation; the reference battery method calculates the SOC of the actual battery stack by designing a branch pipe on the main pipeline and connecting it to a proportionally scaled-down battery stack. However, due to the low internal resistance of the battery, this method is difficult to measure the battery's SOC quickly and accurately during dynamic operation. Furthermore, the precipitation of electrolyte crystals can easily cause blockages in the reference battery's liquid circuit, leading to reference battery failure. With the continuous development of neural network technology, many new applications have emerged in predicting the SOC value of vanadium redox flow battery systems. However, existing methods only consider factors such as voltage, current, and temperature, without considering the impact of charge and discharge time on the SOC prediction of vanadium redox flow batteries. Summary of the Invention

[0004] The following is an overview of the subject matter described in detail herein. This overview is not intended to limit the scope of the claims.

[0005] This application provides a battery SOC prediction method, apparatus, monitoring system, and storage medium, which can more accurately predict the SOC of a battery.

[0006] One embodiment of this application provides a battery SOC prediction method, including: determining the charge / discharge stage of the battery based on the current cumulative charge / discharge duration; selecting a pre-trained SOC prediction model corresponding to the charge / discharge stage of the battery; wherein each charge / discharge stage corresponds to a SOC prediction model; and inputting the current operating parameters of the battery into the corresponding SOC prediction model to obtain the predicted SOC value of the battery.

[0007] One embodiment of this application also provides a battery SOC prediction device, including: a memory and a processor; the memory is used to store a program for battery SOC prediction; the processor is used to read the program for battery SOC prediction and execute the battery SOC prediction method as described in any embodiment of this application.

[0008] One embodiment of this application also provides a battery monitoring system, including: a memory and a processor; the memory is used to store a program for battery SOC prediction; the processor is used to read the program for battery SOC prediction and execute the battery SOC prediction method as described in any embodiment of this application.

[0009] An embodiment of this application also provides a non-transient computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, can implement the battery SOC prediction method as described in any embodiment of this application.

[0010] Compared with related technologies, the battery SOC prediction method, apparatus, monitoring system, and storage medium provided in this application determine the battery's current charge / discharge stage based on its current cumulative charge / discharge duration. Then, a pre-trained SOC prediction model is selected based on the battery's current charge / discharge stage, and the predicted SOC value is obtained based on this model. This solution fully considers the impact of cumulative charge / discharge duration on battery SOC prediction and pre-trains different SOC prediction models for different charge / discharge stages, thus greatly improving the accuracy of SOC value prediction.

[0011] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the application. Other advantages of this application can be realized and obtained by means of the embodiments described in the description and the accompanying drawings. Attached Figure Description

[0012] The accompanying drawings are used to provide an understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0013] Figure 1 This is a simplified flowchart of the battery SOC prediction method according to an embodiment of this application; Figure 2 This is a schematic diagram of the BP neural network structure in an embodiment of this application; Figure 3 This is a flowchart of the battery SOC prediction method according to an embodiment of this application; Figure 4 This is a schematic diagram illustrating the battery SOC prediction model construction process in an embodiment of this application; Figure 5 This is a schematic diagram of a battery SOC prediction device according to an embodiment of this application. Detailed Implementation

[0014] This application describes several embodiments, but these descriptions are exemplary and not limiting, and it will be apparent to those skilled in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are also possible. Unless specifically limited, any feature or element of any embodiment may be used in combination with, or may replace, any feature or element of any other embodiment.

[0015] This application includes and contemplates combinations of features and elements known to those skilled in the art. The embodiments, features, and elements disclosed in this application can also be combined with any conventional features or elements to form unique inventive solutions. Any feature or element of any embodiment can also be combined with features or elements from other inventive solutions to form another unique inventive solution. Therefore, it should be understood that any feature shown and / or discussed in this application can be implemented individually or in any suitable combination. Therefore, the embodiments are not limited except by the limitations imposed by the appended claims and their equivalents. Furthermore, various modifications and changes can be made within the scope of the appended claims.

[0016] Furthermore, in describing representative embodiments, the specification may have presented methods and / or processes as a specific sequence of steps. However, the method or process should not be limited to the specific order of steps described herein, to the extent that it does not depend on such a specific order. As will be understood by those skilled in the art, other sequences of steps are also possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation of the claims. Moreover, the claims concerning the method and / or process should not be limited to the steps performed in the written order, and those skilled in the art will readily understand that these orders can be varied and still remain within the spirit and scope of the embodiments of this application.

[0017] With the continuous development of neural network technology, there are many new applications in the prediction of SOC value of vanadium redox flow battery system. However, existing methods only consider factors such as voltage, current, and temperature, and fail to fully consider the impact of electrolyte flow rate and charge / discharge time on the prediction of SOC of vanadium redox flow battery.

[0018] Therefore, one embodiment of this application provides a battery SOC prediction method, such as... Figure 1 As shown, the following steps may be included: Step S110: Determine the current charge / discharge stage of the battery based on its current cumulative charge / discharge duration; Step S120: Select the corresponding pre-trained SOC prediction model according to the charging and discharging stage of the battery; wherein, each charging and discharging stage corresponds to a SOC prediction model. Step S130: Input the current operating parameters of the battery into the corresponding SOC prediction model to obtain the predicted SOC value of the battery.

[0019] For example, the battery may be a vanadium redox flow battery.

[0020] The battery SOC prediction method in this embodiment determines the battery's current charging / discharging stage based on its current cumulative charging / discharging time. Then, it selects a pre-trained SOC prediction model corresponding to that stage and obtains the predicted SOC value based on that model. This demonstrates that the solution in this embodiment fully considers the impact of cumulative charging / discharging time on battery SOC prediction, and pre-trains different SOC prediction models for different charging / discharging stages, thus significantly improving the accuracy of SOC value prediction.

[0021] In one exemplary embodiment, the SOC prediction model can be trained through the following process: For multiple batteries, multiple first datasets are obtained by following these steps: Battery data throughout its entire lifecycle is collected at preset collection time intervals, and the battery data is preprocessed to obtain charge / discharge data corresponding to different cumulative charge / discharge durations. The entire lifecycle refers to the period from first use to eventual disposal, and the charge / discharge data includes at least the cumulative charge / discharge duration. For example, disposal means the battery can no longer be charged or discharged. The cumulative charge / discharge duration can be recorded based on collection patterns during data acquisition. For instance, since battery data throughout its entire lifecycle is collected at preset collection time intervals, the cumulative charge / discharge duration corresponding to the first collected data is 1 * preset collection time interval, the cumulative charge / discharge duration corresponding to the second collected data is 2 * preset collection time interval, and so on. Alternatively, if the battery management system itself has the function of recording the cumulative charge / discharge duration, the cumulative charge / discharge duration recorded by the battery management system can be directly collected each time data is acquired.

[0022] The battery's charge and discharge data are divided into multiple first datasets based on the cumulative charge and discharge duration range corresponding to each charge and discharge stage; wherein, the first dataset corresponds one-to-one with the charge and discharge stage; the charge and discharge stage is obtained by dividing the battery's entire life cycle according to a preset cumulative duration interval, and each charge and discharge stage corresponds to a cumulative charge and discharge duration range. After collecting the first datasets of multiple batteries, for each charge / discharge stage, a SOC prediction model corresponding to that charge / discharge stage is trained based on the first datasets of multiple batteries corresponding to that charge / discharge stage.

[0023] For example, the preset cumulative duration interval and the preset data collection time interval can be set according to experience or actual needs.

[0024] For example, sensors can be placed at the inlet and outlet of the battery stack to collect data such as voltage, current, inlet and outlet temperatures, and electrolyte flow rate.

[0025] As can be seen from this embodiment, each battery has multiple first datasets, each corresponding to a different charge and discharge stage; after the first datasets of multiple batteries are collected, for each charge and discharge stage, there are multiple first datasets of batteries corresponding to that charge and discharge stage.

[0026] In one example of this embodiment, before dividing the battery's charge and discharge data into multiple first datasets according to the cumulative charge and discharge duration range corresponding to each charge and discharge stage, the method further includes: dividing the battery's entire life cycle into different charge and discharge stages according to a preset cumulative duration interval, and determining the cumulative charge and discharge duration range corresponding to each charge and discharge stage.

[0027] The battery SOC prediction method in this embodiment first divides the entire life cycle of the battery into different charge-discharge stages according to a preset cumulative time interval, and determines the cumulative charge-discharge time range corresponding to each charge-discharge stage. Then, based on the cumulative charge-discharge time range corresponding to each charge-discharge stage, the battery data of the entire life cycle of multiple batteries collected are divided into multiple first datasets (i.e., each first dataset stores the battery data of multiple batteries collected for the charge-discharge stage corresponding to that first dataset). Then, a SOC prediction model is trained for each first dataset to obtain the SOC prediction model corresponding to each charge-discharge stage. In this way, when predicting the SOC value, the corresponding SOC prediction model can be selected for batteries in different charge-discharge stages to predict the SOC value, thereby obtaining a more accurate SOC prediction value.

[0028] In one example of this embodiment, the preprocessing of the battery data to obtain charge / discharge data at different times includes: The battery data is divided into multiple second datasets according to a preset average duration interval; wherein the preset average duration interval is greater than the preset collection time interval and less than the preset cumulative duration interval; For each second dataset, calculate the average of all data in the second dataset from any dimension, and use the average as the new data corresponding to the second dataset; All the new data corresponding to the second dataset are combined sequentially to form the charge / discharge data.

[0029] For example, the preset average duration interval can be set based on experience or actual needs.

[0030] It should be noted that each data object (i.e., each data collection) may include multiple dimensions (such as voltage, current, inlet and outlet temperatures, electrolyte flow rate, each attribute counts as a dimension). The calculation of the average value of all data in the second dataset from any dimension means: for each second dataset, the average value of each dimension of the multiple data objects contained in the second dataset is taken, and finally, a data object composed of the average values ​​of all dimensions is formed, which is the data object corresponding to the second dataset (for distinction, here "a data object composed of the average values ​​of all dimensions" is called the second data object, and "the multiple data objects contained in the original second dataset" is called the first data object; the second data object and the first data object have the same dimensions, and the value of each dimension of the second data object is equal to the average value of the corresponding dimension of the multiple first data objects contained in the original second dataset). The following is a specific example to illustrate this. For instance, a certain second dataset originally includes three data objects: (voltage 2, current 20, temperature 21, internal resistance 20, electrolyte flow rate 300), (voltage 3, current 30, temperature 22, internal resistance 30, electrolyte flow rate 300), and (voltage 4, current 40, temperature 26, internal resistance 40, electrolyte flow rate 400). After processing the data by "calculating the average value of all data in the second dataset from any dimension and using the average value as the new data corresponding to the second dataset", the new data corresponding to the second dataset will be (voltage 3, current 30, temperature 23, internal resistance 30, electrolyte flow rate 300).

[0031] For example, preprocessing the battery data may further include: filtering and cleaning the battery data to remove outliers.

[0032] The battery SOC prediction method in this embodiment, after collecting data at a preset collection time interval, further divides all collected data into multiple parts according to a preset average duration interval. Each part includes (preset average duration interval / preset collection time interval) data points. Then, the average value of each part is calculated, and all the average values ​​are combined to form charge and discharge data. Finally, the SOC prediction model is trained based on the charge and discharge data. This effectively reduces errors in the data collection process, improves the model training effect, and thus improves the accuracy of SOC value prediction.

[0033] The following example illustrates the process of obtaining the first dataset. For instance, the preset cumulative duration interval can be set to 5 days, the preset collection time interval to 1 second, and the preset average duration interval to 1 minute. Assuming the battery's lifespan is 100 days, a total of 100*24*60*60 data points would be collected from the battery's first use until it is scrapped. The process of processing this data to obtain the first dataset can be as follows: The first step is to divide the data into 100*24*60 second datasets at 1-minute intervals (i.e., the preset average duration interval). Then, the average value of the 60 data points in each second dataset is calculated to obtain the average value of the 100*24*60 second datasets. The average value of these 100*24*60 second datasets is used as the charging and discharging data (equivalent to obtaining 100*24*60 charging and discharging data points, with a cumulative charging and discharging time interval of 1 minute for each data point). The second step involves dividing the battery's entire 100-day lifespan into 20 charge-discharge phases, using 5-day intervals (i.e., the preset cumulative duration interval). The cumulative charge-discharge duration for the first charging phase ranges from 1 minute to 5*24*60 minutes, the second charging phase ranges from (5*24*60+1) minutes to 10*24*60 minutes, and so on. The cumulative charge-discharge duration for the 19th charging phase ranges from (90*24*60+1) minutes to 95*24*60 minutes, and the 20th charging phase ranges from (95*24*60+1) minutes to 100*24*60 minutes. Then, the 100*24*60 charge-discharge data points obtained in the first step are divided into 20 first datasets according to the ranges corresponding to each of the above charging-discharge phases.

[0034] In one exemplary embodiment, the charge / discharge data further includes: voltage, current, temperature, internal resistance, and electrolyte flow rate; training the SOC prediction model corresponding to each charge / discharge stage based on the first dataset corresponding to each charge / discharge stage includes: Using the voltage, current, temperature, internal resistance, and electrolyte flow rate as inputs, and the actual SOC value of the battery as a label, a SOC prediction model is trained until the SOC prediction value output by the SOC prediction model meets the preset conditions.

[0035] For example, the phrase "until the SOC prediction value output by the SOC prediction model meets the preset condition" can mean meeting the target accuracy. For instance, when the target accuracy is 1%, the model training is considered complete when the difference between the model's output SOC prediction value and the actual SOC value is controlled within 1%. It should be noted that this application does not limit the specific value of the target accuracy and it can be set according to the actual situation.

[0036] Of course, this application does not limit the "preset condition" in "until the SOC prediction value output by the SOC prediction model meets the preset condition" to "target accuracy", but it can be other conditions.

[0037] The battery SOC prediction method in this embodiment not only considers the influence of voltage, current, temperature, and internal resistance on SOC prediction, but also the influence of electrolyte flow rate on SOC prediction, further improving the accuracy of SOC prediction.

[0038] In one exemplary embodiment, training the SOC prediction model corresponding to each charge / discharge stage based on the first dataset corresponding to each charge / discharge stage includes: Step S210: Based on the first dataset corresponding to the first charge and discharge stage for multiple batteries, train the SOC prediction model corresponding to the first charge and discharge stage. After training is completed, proceed to step S220. Step S220: Based on the first dataset corresponding to the Nth charge / discharge stage for multiple batteries, continue training the SOC prediction model corresponding to the (N-1)th charge / discharge stage that has been trained. After training, determine whether there are any charge / discharge stages for which the corresponding SOC model has not been trained. If so, N = N + 1 and return to step S220; otherwise, end the training. Here, N is a positive integer, and the initial value of N is 2.

[0039] The battery SOC prediction method in this embodiment trains the current SOC prediction model based on the SOC prediction model of the previous charge / discharge stage during the SOC prediction model construction process, and finally iterates to obtain the SOC prediction model for the entire life cycle. In this way, on the one hand, it can save model training time; on the other hand, it can obtain multiple SOC prediction models for different charge / discharge stages, and use the corresponding SOC prediction model for different charge / discharge stages to make predictions, thereby improving the accuracy of SOC prediction.

[0040] In one exemplary embodiment, the SOC prediction model is a BP neural network model. For example... Figure 2As shown, the BP neural network model described in this embodiment can be a 5-N-1 structured BP neural network model. The main structure is an input layer-hidden layer-output layer. Input data starts from the input layer, undergoes a linear combination of weights and biases, is processed by an activation function, and then passed to the hidden layer. The output of each hidden layer then serves as the input to the next layer, and so on, until the output layer produces the final result. Based on the error calculation results, the error signal is propagated back from the output layer to the input layer through the network, continuously correcting the weights and biases to achieve the error target. In this model, the five dimensions of the input layer are the stack internal resistance, voltage, current, temperature, and electrolyte flow rate; the output layer is the SOC value; and the optimal hidden layer nodes are found through continuous model testing.

[0041] Below is a complete example of the battery SOC prediction method of this application, such as... Figure 3 As shown in the diagram (see illustration) Figure 4 This may include the following steps: Step S310: For multiple batteries, collect battery data for the entire battery lifecycle according to a preset data collection time interval. This step can be implemented through the following process: In the vanadium redox flow battery system, corresponding sensors are placed at the inlet and outlet positions of the stack to collect data such as voltage, current, inlet and outlet temperatures, electrolyte flow rate, and cumulative charge and discharge time. Data is collected once every 1 second (i.e., the preset acquisition time interval). At the same time, the internal resistance of each stack is measured before leaving the factory, and the SOC value of each stack is extracted during operation to establish an original database.

[0042] Step S320: Preprocess the battery data to obtain charge / discharge data corresponding to different cumulative charge / discharge durations. This step may specifically include the following steps S321-S323: Step S321: Filter and clean the acquired data, and remove outliers from each group of data, such as data with a value of 0 in one dimension; Step S322: To further eliminate errors in the values ​​of each dimension of the data, average the effective data acquired over 1 minute (i.e., the preset average time interval) for each dimension to obtain new data. Perform the above operation on all data to finally obtain the training and testing data required by the model.

[0043] Step S323: Divide all training data and test data according to the cumulative charge and discharge time. The data generated when the cumulative charge and discharge time reaches 5 days (i.e., the preset cumulative time interval) is a database. Continue until the stack can no longer be charged and discharged. In this way, N databases are obtained in order of charge and discharge time.

[0044] Step S330: For each charge / discharge stage, train the SOC prediction model corresponding to that charge / discharge stage based on the charge / discharge data corresponding to that stage.

[0045] For example, when the preset cumulative duration interval is 5 days, this step may include the following steps S331-S333: Step S331: Input the data from the first database (the database corresponding to the cumulative charge and discharge duration of 1-5 days) into the model, train the model, and end the training when the target accuracy is reached, thus obtaining the real-time SOC prediction model of the vanadium redox flow battery at the initial stage of charge and discharge.

[0046] Step S332: Then, the data in the second database (the database corresponding to the cumulative charge and discharge time of 6-10 days) is input into the first model again for training. Since the mapping relationship in the first model has been established, the internal mapping relationship will be adjusted after the new data is used to train the model, so as to obtain the SOC prediction model for the stage of cumulative charge and discharge time of 6-10 days.

[0047] Step S333: Input the data from the third database (the database corresponding to the cumulative charge / discharge duration of 11-15 days) into the second model to obtain the SOC prediction model for the stage of 11-15 days of cumulative charge / discharge time. Repeat this process to finally obtain the SOC prediction model for each charge / discharge stage of the vanadium redox flow battery.

[0048] Step S340: Integrate multiple trained SOC prediction models into the battery monitoring system. When performing SOC prediction, determine the charging and discharging stage of the battery based on the current cumulative charging and discharging time. Then, select the corresponding pre-trained SOC prediction model based on the charging and discharging stage of the battery. Finally, input the current operating parameters of the battery into the corresponding SOC prediction model to obtain the SOC prediction value of the battery.

[0049] In step S340, the SOC prediction models for different charge and discharge stages are integrated into the vanadium redox flow system. Then, the battery voltage, stack, temperature, internal resistance, and electrolyte flow rate are input into the corresponding SOC prediction model in the system, and the predicted value of the battery's SOC can be obtained in real time.

[0050] In summary, this embodiment leverages the powerful nonlinear processing capabilities of the BP neural network to accurately and rapidly predict the State of Charge (SOC) of a vanadium redox flow battery in real time. Furthermore, considering the impact of charge and discharge time on SOC, the model is constructed segmented according to charge and discharge time, resulting in a set of SOC prediction models covering the entire lifecycle of the vanadium redox flow battery. This allows for accurate prediction of the SOC value of the battery at different charge and discharge stages.

[0051] One embodiment of this application provides a battery SOC prediction device, such as... Figure 5 As shown, it includes: memory and processor; The memory is used to store the program for predicting battery SOC; The processor is configured to read the program for battery SOC prediction and execute the battery SOC prediction method as described in any embodiment of this application.

[0052] An embodiment of this application also provides a battery monitoring system, including: a memory and a processor; the memory is used to store a program for predicting battery SOC; the processor is used to read the program for predicting battery SOC and execute the battery SOC prediction method as described in any embodiment of this application.

[0053] An embodiment of this application also provides a non-transient computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, can implement the battery SOC prediction method as described in any embodiment of this application.

[0054] In summary, the battery SOC prediction method, device, monitoring system, and storage medium provided in this application have the following beneficial effects: (1) Compared with existing SOC prediction methods, it comprehensively considers factors such as voltage, current, in-pile and out-of-pile temperature, charge and discharge time, and electrolyte flow rate, thereby improving the accuracy of SOC prediction. (2) In the process of constructing the SOC prediction model, the charge and discharge time is divided into fixed time intervals, and relevant data are collected. Based on the previous SOC prediction model, the SOC prediction model is further trained by combining the data of the current time period to obtain a new SOC prediction model. Finally, the SOC prediction model for the entire life cycle is obtained through continuous iteration. In this way, on the one hand, the training time of the model can be saved; on the other hand, multiple SOC prediction models for different charge and discharge stages can be obtained. The corresponding SOC prediction model is used for prediction for different charge and discharge stages, thereby improving the accuracy of SOC prediction. (3) In terms of data processing, the average value of the data obtained at the preset average time interval can effectively reduce the error in the data acquisition process, thereby improving the training effect of the SOC prediction model and thus improving the accuracy of SOC prediction.

[0055] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term "computer storage medium" includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0056] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0057] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for predicting battery SOC, characterized in that, include: The charging / discharging stage of the battery is determined based on the current cumulative charging / discharging time. The corresponding pre-trained SOC prediction model is selected based on the charging and discharging stage of the battery; each charging and discharging stage corresponds to a separate SOC prediction model. Input the battery’s current operating parameters into the corresponding SOC prediction model to obtain the battery’s predicted SOC value. The SOC prediction model is trained through the following process: For multiple batteries, multiple first datasets are obtained according to the following steps: Battery data for the entire life cycle of the battery is collected at preset collection time intervals, and the battery data is preprocessed to obtain charge and discharge data corresponding to different cumulative charge and discharge times. The entire life cycle refers to the entire period from first use to scrapping, and the charge and discharge data includes at least the cumulative charge and discharge time. The charge and discharge data of the battery is divided into multiple first datasets according to the cumulative charge and discharge time range corresponding to each charge and discharge stage. The first dataset corresponds one-to-one with the charge and discharge stage, and the charge and discharge stage is obtained by dividing the entire life cycle of the battery according to the preset cumulative time interval. Each charge and discharge stage corresponds to a cumulative charge and discharge time range. After collecting the first datasets of multiple batteries, for each charge / discharge stage, a SOC prediction model corresponding to that charge / discharge stage is trained based on the first datasets of multiple batteries corresponding to that charge / discharge stage.

2. The battery SOC prediction method according to claim 1, characterized in that, The preprocessing of the battery data to obtain charge / discharge data corresponding to different cumulative charge / discharge durations includes: The battery data is divided into multiple second datasets according to a preset average duration interval; wherein the preset average duration interval is greater than the preset collection time interval and less than the preset cumulative duration interval; For each second dataset, calculate the average of all data in the second dataset from any dimension, and use the average as the new data corresponding to the second dataset; All the new data corresponding to the second dataset are combined sequentially to form the charge / discharge data.

3. The battery SOC prediction method according to claim 1, characterized in that, The charge / discharge data also includes: voltage, current, temperature, internal resistance, and electrolyte flow rate; training the SOC prediction model corresponding to each charge / discharge stage based on the first dataset for each stage includes: Using the voltage, current, temperature, internal resistance, and electrolyte flow rate as inputs, and the actual SOC value of the battery as a label, a SOC prediction model is trained until the SOC prediction value output by the SOC prediction model meets the preset conditions.

4. The battery SOC prediction method according to claim 1, characterized in that, The step of training a SOC prediction model corresponding to the charge / discharge stage based on a first dataset of multiple batteries corresponding to the charge / discharge stage includes:

51. Based on the first dataset corresponding to the first charge and discharge stage for multiple batteries, train the SOC prediction model corresponding to the first charge and discharge stage. After completing the training, proceed to step 52.

52. Based on the first dataset corresponding to the Nth charge / discharge stage for multiple batteries, continue training the SOC prediction model corresponding to the N-1th charge / discharge stage that has been trained. After training, determine whether there are any charge / discharge stages for which the corresponding SOC model has not been trained. If so, N=N+1 and return to step 52; otherwise, end the training. The initial value of N is 2.

5. The battery SOC prediction method according to claim 1, characterized in that: The SOC prediction model is a BP neural network model.

6. The battery SOC prediction method according to claim 1, characterized in that: The battery is a vanadium redox flow battery.

7. A battery SOC prediction device, comprising: The memory and processor are characterized by: The memory is used to store the program for predicting battery SOC; The processor is configured to read the program for battery SOC prediction and execute the battery SOC prediction method as described in any one of claims 1 to 6.

8. A battery monitoring system, comprising: The memory and processor are characterized by: The memory is used to store the program for predicting battery SOC; The processor is configured to read the program for battery SOC prediction and execute the battery SOC prediction method as described in any one of claims 1 to 6.

9. A non-transient computer-readable storage medium storing a computer program, wherein, When the computer program is executed by a processor, it can implement the battery SOC prediction method as described in any one of claims 1 to 6.

Images