Battery state prediction method and device and electronic equipment

Abstract

The invention discloses a battery state prediction method and device and electronic equipment. The method relates to the field of artificial intelligence, and comprises the following steps: obtaining operation data of a target battery in a first historical time period, and obtaining N groups of first historical operation data; inputting the N groups of first historical operation data into an operation data prediction model to obtain an operation data set of the target battery at H target moments in the target time period; m operation data sequences of the target battery in the target time period are generated; under the condition that an abnormal fluctuation index exists in the fluctuation indexes of the M operation data sequences, determining that the target battery is abnormal; and under the condition that no abnormal fluctuation index exists in the fluctuation indexes of the M operation data sequences, obtaining an initial residual capacity value of the target battery in the first historical time period, and determining the operation state of the target battery according to the initial residual capacity value. Through the method and the device, the problem of low accuracy of determining the battery state according to the use duration in related technologies is solved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence, and more specifically, to a method, apparatus, and electronic device for predicting battery state. Background Technology

[0002] Currently, financial institutions typically set up backup power systems to ensure the normal operation of data centers, enabling seamless switching during mains power outages and maintaining the continuous operation of critical data center businesses. Among these backup power systems, batteries are a core component, and their performance and reliability directly affect the overall stability and security of the data center.

[0003] To ensure the backup power system operates normally, batteries need to be maintained and replaced regularly. However, the maintenance and replacement strategy mainly relies on the battery's age to determine its condition and whether to replace it. While this strategy ensures that all batteries in the backup power system are in normal operating condition, it ignores the performance differences between different manufacturers, models, and individual batteries. Due to these differences, the health status of batteries with the same age can vary significantly. Replacing batteries according to a fixed age not only fails to effectively reflect their actual condition but also results in a huge waste of resources and increases the operating costs of the data center.

[0004] There is currently no effective solution to the problem of low accuracy in determining battery status based on usage time in related technologies. Summary of the Invention

[0005] The main objective of this application is to provide a method, apparatus, and electronic device for predicting battery status, in order to solve the problem of low accuracy in determining battery status based on usage time in related technologies.

[0006] To achieve the above objectives, according to one aspect of this application, a method for predicting battery state is provided. The method includes: acquiring operational data of a target battery at different historical moments within a first historical period, obtaining N sets of first historical operational data, wherein each set of first historical operational data includes M operational data points of the target battery at the corresponding historical moment, where M is the number of types of indicator data for the target battery, and M and N are positive integers; inputting the N sets of first historical operational data into an operational data prediction model to obtain a set of operational data for the target battery at H target moments within a target period, resulting in H sets of operational data, where H is a positive integer; generating M operational data sequences of the target battery within the target period based on the H sets of operational data; determining that the target battery is abnormal if abnormal fluctuation indicators exist among the fluctuation indicators of the M sets of operational data; and obtaining the initial remaining capacity value of the target battery within the first historical period if no abnormal fluctuation indicators exist among the fluctuation indicators of the M sets of operational data, and determining the operational state of the target battery based on the initial remaining capacity value.

[0007] Optionally, the operational data prediction model is trained as follows: Operational data of the target battery within P second historical time periods are acquired, resulting in P sets of second historical operational data, where P is a positive integer. Each set of second historical operational data includes Q sets of second historical operational data within the second historical time period, where Q is a positive integer. Third historical time periods adjacent to each second historical time period are acquired, where each third historical time period is later than the adjacent second historical time period, and operational data within each third historical time period are acquired, resulting in P sets of third historical operational data. The second and third historical operational data sets with adjacent relationships are combined into a set of sample data, resulting in P sets of first sample data. The initial prediction model is trained using the P sets of first sample data to obtain the operational data prediction model, where the initial prediction model is a neural network model with multiple output layers.

[0008] Optionally, the method further includes: for any running data sequence, calculating the dispersion of the running data sequence and obtaining the dispersion threshold of the running data sequence, wherein the dispersion is at least one of the following: variance, standard deviation; comparing the dispersion with the dispersion threshold, and if the dispersion is greater than the dispersion threshold, determining the fluctuation index of the running data sequence as an abnormal fluctuation index; if the dispersion is less than or equal to the dispersion threshold, determining the anomaly identification result of the running data sequence based on the values ​​of the running data in the running data sequence.

[0009] Optionally, determining the anomaly identification result of the running data sequence based on the values ​​of the running data in the running data sequence includes: obtaining the data threshold range of the running data sequence and determining whether there is abnormal data outside the data threshold range in the running data sequence; if there is abnormal data, determining the fluctuation index of the running data sequence as an abnormal fluctuation index; if there is no abnormal data, determining the fluctuation index of the running data sequence as a normal fluctuation index.

[0010] Optionally, obtaining the initial remaining capacity value of the target battery in the first historical period includes: obtaining the initial remaining capacity value of the target battery at each historical moment in the first historical period to obtain N initial remaining capacity values; and calculating the average of the N initial remaining capacity values ​​to obtain the initial remaining capacity value.

[0011] Optionally, determining the operating status of the target battery based on the initial remaining capacity value includes: obtaining the nominal capacity of the target battery, dividing the initial remaining capacity value by the nominal capacity to obtain the health status index value of the target battery; determining whether the health status index value is less than the index threshold; if the health status index value is less than the index threshold, determining that the operating status of the target battery is abnormal; if the health status index value is greater than or equal to the index threshold, determining that the operating status of the target battery is normal.

[0012] Optionally, the method further includes: determining the cause of the target battery's abnormality when the target battery is determined to be abnormal; generating alarm information and maintenance information for maintaining the target battery based on the cause of the abnormality; and sending the alarm information and maintenance information to the operation and maintenance terminal that is related to the cause of the abnormality.

[0013] To achieve the above objectives, according to another aspect of this application, a battery state prediction device is provided. The device includes: a first acquisition unit, configured to acquire operating data of a target battery at different historical moments within a first historical period, obtaining N sets of first historical operating data, wherein each set of first historical operating data includes M operating data points of the target battery at the corresponding historical moment, where M is the number of types of indicator data for the target battery, and M and N are positive integers; a prediction unit, configured to input the N sets of first historical operating data into an operating data prediction model to obtain a set of operating data of the target battery at H target moments within a target period, obtaining H sets of operating data, where H is a positive integer; a first generation unit, configured to generate M operating data sequences of the target battery within the target period based on the H sets of operating data; a first determination unit, configured to determine that the target battery is abnormal if abnormal fluctuation indicators exist among the fluctuation indicators of the M sets of operating data; and a second determination unit, configured to acquire the initial remaining capacity value of the target battery within the first historical period and determine the operating state of the target battery based on the initial remaining capacity value if no abnormal fluctuation indicators exist among the fluctuation indicators of the M sets of operating data.

[0014] Optionally, the operational data prediction model is trained as follows: a second acquisition unit is used to acquire operational data of the target battery within P second historical time periods, resulting in P sets of second historical operational data, where P is a positive integer, and each set of second historical operational data includes Q sets of second historical operational data within the second historical time period, where Q is a positive integer; a third acquisition unit is used to acquire third historical time periods adjacent to each second historical time period, where each third historical time period is later than the adjacent second historical time period, and acquire operational data within each third historical time period, resulting in P sets of third historical operational data; a combination unit is used to combine the second historical operational data sets and the third historical operational data sets with adjacent relationships into a set of sample data, resulting in P sets of first sample data; and a training unit is used to train the initial prediction model using the P sets of first sample data to obtain the operational data prediction model, wherein the initial prediction model is a neural network model with multiple output layers.

[0015] Optionally, the device further includes: a calculation unit, used to calculate the dispersion of any running data sequence and obtain a dispersion threshold for the running data sequence, wherein the dispersion is at least one of the following: variance, standard deviation; a comparison unit, used to compare the dispersion with the dispersion threshold, and determine the fluctuation index of the running data sequence as an abnormal fluctuation index if the dispersion is greater than the dispersion threshold; and a third determination unit, used to determine the anomaly identification result of the running data sequence based on the value of the running data in the running data sequence if the dispersion is less than or equal to the dispersion threshold.

[0016] Optionally, the third determining unit includes: a first judging module, used to obtain the data threshold range of the running data sequence and judge whether there is abnormal data outside the data threshold range in the running data sequence; a first determining module, used to determine the fluctuation index of the running data sequence as an abnormal fluctuation index when there is abnormal data; and a second determining module, used to determine the fluctuation index of the running data sequence as a normal fluctuation index when there is no abnormal data.

[0017] Optionally, the second determining unit includes: a first acquisition module, used to acquire the initial remaining capacity value of the target battery at each historical moment within the first historical period, to obtain N initial remaining capacity values; and a calculation module, used to calculate the average of the N initial remaining capacity values ​​to obtain the initial remaining capacity value.

[0018] Optionally, the second determining unit includes: a second acquiring module, used to acquire the nominal capacity of the target battery and divide the initial remaining capacity value by the nominal capacity to obtain the health status index value of the target battery; a second judging module, used to judge whether the health status index value is less than the index threshold; a third determining module, used to determine that the operating state of the target battery is abnormal when the health status index value is less than the index threshold; and a fourth determining module, used to determine that the operating state of the target battery is normal when the health status index value is greater than or equal to the index threshold.

[0019] Optionally, the device further includes: a fourth determining unit, used to determine the cause of the target battery's abnormality when the target battery is determined to be abnormal; and a second generating unit, used to generate alarm information and maintenance information for maintaining the target battery based on the cause of the abnormality, and send the alarm information and maintenance information to the operation and maintenance terminal that is related to the cause of the abnormality.

[0020] To achieve the above objectives, according to another aspect of this application, an electronic device is provided, the electronic device including a memory storing an executable program; and a processor for running the program, wherein the program executes the above-described battery state prediction method during runtime.

[0021] To achieve the above objectives, according to another aspect of this application, a computer program product is provided, including computer instructions that, when executed by a processor, implement the steps of the above-described battery state prediction method.

[0022] In this embodiment, the operation data of the target battery at different historical moments within a first historical period are obtained to form N sets of first historical operation data. Each set of first historical operation data includes M operation data points of the target battery at the corresponding historical moment, where M is the number of types of indicator data of the target battery, and M and N are positive integers. The N sets of first historical operation data are input into the operation data prediction model to obtain H sets of operation data of the target battery at target moments within the target period, where H is a positive integer. Based on the H sets of operation data, M operation data sequences of the target battery within the target period are generated. Abnormal fluctuation indicators exist in the fluctuation indicators of the M operation data sequences. In the case of a target battery, an anomaly is identified. If no abnormal fluctuation indicators are found among the fluctuation indicators of M operational data sequences, the initial remaining capacity value of the target battery in the first historical period is obtained, and the operational status of the target battery is determined based on the initial remaining capacity value. The operational data under the target period is predicted by a deep learning prediction model, and then the operational data sequence of various indicator data is determined based on the operational data. By analyzing the abnormal fluctuations of the operational data sequence, the goal of accurately identifying the battery status is achieved. This realizes the technical effect of optimizing the battery replacement strategy, reducing resource waste, and lowering costs, thereby solving the technical problem of low accuracy in determining battery status based on usage time in related technologies. Attached Figure Description

[0023] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0024] Figure 1 A hardware block diagram of a computer terminal for implementing a method for predicting battery state is shown.

[0025] Figure 2 This is a flowchart of a battery state prediction method according to Embodiment 1 of this application;

[0026] Figure 3 This is a schematic diagram of an optional long short-term memory neural network model provided according to Embodiment 1 of this application;

[0027] Figure 4 This is a schematic diagram of a battery state prediction device according to Embodiment 2 of this application;

[0028] Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0029] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0030] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0032] It should be noted that the battery state prediction method, apparatus, and electronic device determined in this disclosure can be used in the field of artificial intelligence, or in any field other than artificial intelligence, such as the field of fintech. The application fields of the battery state prediction method, apparatus, and electronic device determined in this disclosure are not limited.

[0033] It should be noted that all information, user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, and displayed data) used in this application are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with the relevant laws, regulations, and standards of the relevant regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse use. If the user chooses to refuse, the process proceeds to the expert decision-making process. For example, this system has interfaces with relevant users or organizations. Before obtaining relevant information, a request to obtain the information needs to be sent to the aforementioned user or organization through the interface, and the relevant information is obtained only after receiving consent from the aforementioned user or organization.

[0034] The embodiments or examples disclosed herein are not exhaustive, but merely illustrative of some embodiments or examples, and are not intended to limit the scope of protection of this disclosure. Unless otherwise specified, each step in a particular embodiment or example can be implemented as an independent embodiment, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a particular embodiment or example can also be implemented as an independent embodiment, and the order of the steps in a particular embodiment or example can be arbitrarily interchanged. Furthermore, optional methods or examples in a particular embodiment or example can be arbitrarily combined; moreover, embodiments or examples can be arbitrarily combined. For example, some or all steps of different embodiments or examples can be arbitrarily combined, and a particular embodiment or example can be arbitrarily combined with optional methods or examples of other embodiments or examples.

[0035] For ease of description, the following explains some of the nouns or terms used in the embodiments of this application:

[0036] Long Short-Term Memory (LSTM) is a type of temporal recurrent neural network that is well-suited for processing and predicting time series problems with very long intervals and delays.

[0037] Example 1

[0038] According to an embodiment of this application, an embodiment of a method for predicting battery state is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0039] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal for implementing a method for predicting battery state is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0040] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0041] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the battery state prediction method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned battery state prediction method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0042] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0043] The display may be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0044] Under the aforementioned operating environment, this application provides the following: Figure 2The method for predicting battery state is shown. Figure 2 This is a flowchart of the battery state prediction method provided in Embodiment 1 of this application, as follows: Figure 2 As shown, the method includes:

[0045] Step S201: Obtain the operating data of the target battery at different historical moments within the first historical period to obtain N sets of first historical operating data. Each set of first historical operating data includes M operating data of the target battery at the corresponding historical moment, where M is the number of types of indicator data of the target battery, and M and N are positive integers.

[0046] It should be noted that the target battery refers to the specific battery for which health status prediction is to be performed. The operating data at different historical moments refers to the operating status data of the battery at each historical moment, including but not limited to charge and discharge capacity, charge and discharge current, battery voltage, current, temperature, etc. N is N historical time periods, and M is M index data. Each set of first historical operating data includes the index values ​​of M index data at that historical moment, thus obtaining N sets of first historical operating data.

[0047] Specifically, when predicting the operating status of a target battery, it is necessary to first obtain the basic data for prediction, namely the operating data of the target battery in the first historical period. In other words, multiple operating indicators of the battery are recorded regularly in the first historical period, including charge and discharge capacity, charge and discharge current, battery voltage, current, temperature and other environmental factors that may affect battery life. Then, the operating data of the target battery in future periods can be predicted by obtaining the historical operating data, and the status of the target battery can be determined based on the predicted data.

[0048] Step S202: Input N sets of first historical operating data into the operating data prediction model to obtain the operating data set of the target battery at H target times within the target time period, resulting in H operating data sets, where H is a positive integer.

[0049] It should be noted that the operational data prediction model refers to a deep learning model trained on historical data that can predict future operational data, and the target period refers to the time period that needs to be predicted in the future, such as the next 10 days.

[0050] Specifically, after obtaining N sets of first historical operating data, the N sets of first historical operating data can be input into the operating data prediction model. The operating data prediction model predicts the operating data of various indicators of the target battery at each target time (i.e., future time), resulting in H sets of operating data. Each set of operating data includes multiple indicator data of the target battery at the target time corresponding to that set, such as battery voltage, current, temperature, etc.

[0051] Step S203: Generate M operational data sequences for the target battery within the target time period based on H operational data sets.

[0052] Specifically, after obtaining H sets of operational data, since each set of operational data includes M indicator data of the target battery at a target time, the H sets of operational data can be changed from being classified by time to being classified by indicator, thereby obtaining operational data of M indicator data at multiple target times. The operational data is then sorted by time to obtain M operational data sequences, and the fluctuation of each indicator data and the indicator value of each indicator data can be determined based on the operational data sequences.

[0053] Step S204: If there are abnormal fluctuation indicators among the fluctuation indicators of the M running data sequences, the target battery is determined to be abnormal.

[0054] Specifically, after obtaining M operational data sequences, fluctuation analysis needs to be performed on these sequences to identify indicators of abnormal fluctuations. This can be done by calculating the standard deviation, maximum-minimum difference, or by using machine learning algorithms (such as Isolation Forest, Autoencoder, etc.) to detect outliers. If any fluctuation in any data sequence exceeds the preset normal range, the target battery is considered to potentially be abnormal.

[0055] By identifying abnormal fluctuation indicators, it is possible to detect sudden deterioration in battery performance in a timely manner, take necessary preventive measures or replace the battery in advance, and avoid power supply interruptions caused by battery failure.

[0056] Step S205: If there are no abnormal fluctuation indicators among the fluctuation indicators of the M running data sequences, obtain the initial remaining capacity value of the target battery in the first historical period, and determine the operating status of the target battery based on the initial remaining capacity value.

[0057] Specifically, when there are no abnormal fluctuation indicators, all indicator data are in a stable state. At this time, the change in the remaining capacity of the battery can be further calculated, and the operating status of the target battery can be determined based on the initial remaining capacity value. Then, the operating status of the battery can be determined by both the fluctuation of indicator values ​​and the remaining capacity value, thereby ensuring the accuracy of determining the operating status of the battery.

[0058] The battery state prediction method provided in this application embodiment obtains N sets of first historical operating data by acquiring the operating data of the target battery at different historical moments within a first historical period. Each set of first historical operating data includes M operating data points of the target battery at the corresponding historical moment, where M is the number of types of indicator data of the target battery, and M and N are positive integers. The N sets of first historical operating data are input into an operating data prediction model to obtain H sets of operating data of the target battery at target moments within a target period, where H is a positive integer. Based on the H sets of operating data, M operating data sequences of the target battery within the target period are generated. The fluctuation indicators of the M operating data sequences are stored... In the case of abnormal fluctuation indicators, the target battery is identified as abnormal. If no abnormal fluctuation indicators are found among the fluctuation indicators of M operating data sequences, the initial remaining capacity value of the target battery in the first historical period is obtained, and the operating status of the target battery is determined based on the initial remaining capacity value. The operating data under the target period is predicted by a deep learning prediction model, and then the operating data sequence of various indicator data is determined based on the operating data. By analyzing the abnormal fluctuations of the operating data sequence, the goal of accurately identifying the battery status is achieved. This realizes the technical effect of optimizing the battery replacement strategy, reducing resource waste, and lowering costs, thereby solving the technical problem of low accuracy in determining the battery status based on usage time in related technologies.

[0059] To ensure the accuracy of the model's prediction results, optionally, in the battery state prediction method provided in this application embodiment, the operating data prediction model is trained in the following manner: The operating data of the target battery within P second historical time periods is obtained, resulting in P sets of second historical operating data, where P is a positive integer, and each set of second historical operating data includes Q sets of second historical operating data within the second historical time period, where Q is a positive integer; A third historical time period adjacent to each second historical time period is obtained, where each third historical time period is later than the adjacent second historical time period, and operating data within each third historical time period is obtained, resulting in P sets of third historical operating data; The second and third historical operating data sets with adjacent relationships are combined into a set of sample data, resulting in P sets of first sample data; The initial prediction model is trained using the P sets of first sample data to obtain the operating data prediction model, wherein the initial prediction model is a neural network model with multiple output layers.

[0060] It should be noted that the initial prediction model used in this embodiment can be a Long Short-Term Memory Neural Network (LSTM) model. Figure 3 This is a schematic diagram of an optional long short-term memory neural network model provided in Embodiment 1 of this application, as shown below. Figure 3As shown, LSTM contains multiple cells for processing data and generating output results. Each cell consists of four neural networks that interact in a special way. The output of one node is used as the input of other nodes, and each line passes a complete vector.

[0061] The key to LSTM is the state of the cell, i.e. Figure 3 The top horizontal line, the cell state is like a conveyor belt, and only some minor linear interactions are needed to pass through the entire chain structure, which is the place where information is stored. Therefore, information can easily flow through it in an invariable form.

[0062] To add or remove information from a cell, LSTM uses control gates. These gates determine how information passes through, consisting of a sigmoid neural network layer and a pointwise multiplication operation. The sigmoid layer outputs a number between 0 and 1, and the multiplication operation determines how much information can be transmitted: 0 means no information is transmitted, and 1 means all information is transmitted. LSTM uses three such control gates to protect and control the cell state.

[0063] To add / remove information from a cell, LSTM uses three gates: a forget gate, an input gate, and an output gate, to protect and control the cell state. Specifically:

[0064] Forget Gate: Determines which information to delete. For inputs xt and ht-1, the forget gate outputs a number in the range [0, 1], which is placed into the cell state Ct-1. When the value is 0, all information is deleted; when the value is 1, all information is retained. The value between 0 and 1 determines the weight of the information to be retained.

[0065] Input gate: determines which information should be retained. Secondly, it uses a tanh layer to generate a vector C~t for this information, which is used to update the cell state.

[0066] Output gate: Determines the cell content to be output, then uses a tanh layer to push the cell state value between -1 and 1, and multiplies it by the output of the sigmoid layer, so as to output only the part that is desired.

[0067] The influence of the model on the model is controlled by a neural network through forget gates, input gates, and output gates. Finally, the model is optimized and demonstrated through methods such as cross-validation.

[0068] It should be noted that the output gate of the initial prediction model in this embodiment is a multi-output layer. Unlike the single-output model, the multi-output model has one or more output layers in the last few layers of the network. Each output layer is responsible for predicting a specific target or label, thus obtaining output data at different target time points based on a single input.

[0069] Specifically, when training a neural network model, it is first necessary to acquire a second historical operating data set and a corresponding third historical operating data set. The second historical period refers to a specific past time segment used for model training; these time segments can be continuous or intermittent, but they are all earlier than the first historical period (the period used for direct prediction) to obtain more training data. The third historical period refers to the time segment immediately following the second historical period, used to collect recent battery operating status data so that the model can learn the continuity of battery performance changes over time based on the second and third historical operating data sets.

[0070] After obtaining the data required for training, each second historical running data set and the corresponding third historical running data set can be combined into a set of sample data to obtain multiple sets of first sample data. The model can then be trained using each set of sample data, enabling the model to predict the corresponding third historical running data set based on the second historical running data set. This ensures that the running data prediction model after training can predict the running data at each moment in the target time period based on the first historical running data in the historical time period.

[0071] This embodiment trains a high-precision operational data prediction model using historical operational data. This model can predict the future state of multiple indicators based on the continuity and immediate impact of battery performance changes at historical moments, significantly improving the accuracy and practicality of the prediction.

[0072] For example, in practical applications, the model can predict that the battery's charge and discharge capacity will drop to a certain level in the next quarter, the temperature will show an abnormal upward trend, and the voltage will remain stable. This allows data centers to plan battery maintenance and replacement in advance, avoid power supply interruptions caused by battery failures, effectively control operation and maintenance costs, and improve overall operating efficiency and safety.

[0073] To accurately identify whether there are anomalies in the running data sequence, optionally, in the battery state prediction method provided in this application embodiment, the method further includes: for any running data sequence, calculating the dispersion of the running data sequence and obtaining the dispersion threshold of the running data sequence, wherein the dispersion is at least one of the following: variance, standard deviation; comparing the dispersion with the dispersion threshold, and if the dispersion is greater than the dispersion threshold, determining the fluctuation index of the running data sequence as an abnormal fluctuation index; if the dispersion is less than or equal to the dispersion threshold, determining the anomaly identification result of the running data sequence based on the value of the running data in the running data sequence.

[0074] It should be noted that dispersion refers to the degree of dispersion of data distribution. Commonly used indicators include variance and standard deviation, which can reflect the magnitude of fluctuation of data points relative to the mean. The dispersion threshold is a preset value used as a benchmark to judge whether the data sequence has abnormal fluctuations. If the dispersion of the sequence exceeds this threshold, the sequence is generally considered to have abnormal fluctuations.

[0075] Specifically, when determining whether the running data sequence is fluctuating abnormally, the degree of fluctuation can be determined by calculating the dispersion of the running data sequence. When multiple indicator data running data sequences are obtained, the running data sequences of each indicator data can be analyzed separately. For any running data sequence, the variance or standard deviation of the sequence can be calculated, and the variance threshold or standard deviation threshold can be obtained.

[0076] Next, the variance is compared with the variance threshold, or the standard deviation is compared with the standard deviation threshold. If the dispersion is greater than the threshold, the volatility index of the sequence is considered to be an abnormal volatility index. Conversely, if the dispersion is less than or equal to the threshold, the sequence is further checked for numerical anomalies. Then, by comparing the dispersion with the threshold, the running data sequences with abnormal fluctuations are screened out.

[0077] When the dispersion is greater than the threshold, the fluctuation index of the sequence can be directly identified as an abnormal fluctuation index. When the dispersion is less than or equal to the threshold, a deeper numerical analysis is required, that is, to detect extreme values ​​in the sequence, to determine whether there are values ​​that exceed the normal range, and to further confirm whether there are other types of anomalies in the battery. By identifying the values ​​in the sequence, it is possible to determine whether there are anomalies in the running data sequence, thus ensuring the accuracy of anomaly identification in the sequence.

[0078] For example, suppose we obtain the daily charging capacity sequence of a target battery for the next week, with the data as follows: {75, 76, 78, 79, 80, 82, 85} mAh, and an average value of approximately 79 mAh. The standard deviation of this sequence is calculated to be 3 mAh. Under normal operating conditions, the standard deviation of charging capacity should typically be less than 2 mAh; therefore, the preset dispersion threshold is 2 mAh. Comparing the calculated standard deviation of 3 mAh with the threshold of 2 mAh reveals that the standard deviation is greater than the threshold. This indicates abnormal fluctuations in the battery's charging capacity, which may suggest instability in the ongoing chemical reaction within the battery or the presence of external influencing factors. Further investigation of the battery's health is needed. Therefore, by identifying and analyzing the fluctuations in the data sequence, we can complete the identification of the battery's operating status.

[0079] This embodiment can promptly capture abnormal fluctuations in battery performance indicators by calculating the dispersion and comparing it with a preset threshold. Furthermore, by combining the numerical anomaly analysis of the data sequence, it can still identify potential health risks of the battery even when the dispersion is relatively stable. This improves the intelligence level of battery health management, helps data center managers make more reasonable and timely maintenance decisions, thereby extending battery life, reducing operating costs, and improving the overall reliability and safety of the system.

[0080] To accurately identify whether there are anomalies in the operating data sequence, optionally, in the battery state prediction method provided in this application embodiment, determining the anomaly identification result of the operating data sequence based on the values ​​of the operating data in the operating data sequence includes: obtaining the data threshold range of the operating data sequence, and determining whether there is abnormal data in the operating data sequence that is outside the data threshold range; if there is abnormal data, determining the fluctuation index of the operating data sequence as an abnormal fluctuation index; if there is no abnormal data, determining the fluctuation index of the operating data sequence as a normal fluctuation index.

[0081] Specifically, when the dispersion is less than or equal to the threshold, it is necessary to determine whether there is abnormal data in the numerical sequence. In this case, the data threshold range of the running data sequence can be obtained, and it can be determined whether there is abnormal data outside the range. For example, for the running data sequence of battery temperature, the data threshold range can be between 20°C and 30°C, and it can be determined whether there are values ​​in the temperature data sequence that exceed the threshold range.

[0082] If abnormal data is found in the running data sequence, the sequence will be marked as having abnormal fluctuation indicators even if the overall dispersion of the sequence does not exceed the standard. Conversely, if no abnormal data is found, even if the dispersion is slightly large, the fluctuation indicators of the sequence will be regarded as normal fluctuation indicators. This ensures the comprehensiveness of the detection process, which not only considers the degree of fluctuation of the data sequence, but also analyzes the specific values ​​of the data in depth, preventing misjudgments of battery status due to a small number of extreme values.

[0083] For example, given a battery voltage operating data sequence for the next month, with values ​​of {3.5V, 3.6V, 3.7V, 3.8V, 3.9V, 4.2V}, historical data analysis determined the normal voltage fluctuation range to be 3.3V to 4.0V. Upon examining this sequence, a voltage reading of 4.2V on day 6 was found, significantly exceeding the threshold range. Although the overall voltage fluctuation was within acceptable limits, the presence of this abnormal data point led to the determination of this operating data sequence as an abnormal fluctuation indicator, thus ensuring accurate identification of abnormal operating data sequences.

[0084] This embodiment achieves the technical effect of improving the accuracy of identifying abnormal operating data sequences by combining the detection of numerical anomalies in the running data sequence with the comprehensive judgment of fluctuation indicators.

[0085] To accurately calculate the initial remaining capacity value, optionally, in the battery state prediction method provided in the embodiments of this application, obtaining the initial remaining capacity value of the target battery in the first historical period includes: obtaining the initial remaining capacity value of the target battery at each historical moment in the first historical period to obtain N initial remaining capacity values; calculating the average of the N initial remaining capacity values ​​to obtain the initial remaining capacity value.

[0086] It should be noted that the initial remaining capacity value is the capacity that the battery can release when fully charged, i.e., the capacity when fully charged.

[0087] Specifically, when calculating the initial remaining capacity value, the initial remaining capacity value of the target battery at each historical moment within the first historical period can be obtained. That is, the value obtained after measuring the initial remaining capacity of the target battery at each historical moment, and the average value of N initial remaining capacity values ​​is calculated to obtain the initial remaining capacity value of the target battery within the historical period. Then, based on the initial remaining capacity value, the average health status and storage capacity of the target battery within the first historical period can be determined, thereby determining whether there is any abnormality in the target battery.

[0088] For example, suppose the remaining capacity of the target battery is measured at the end of each month during the first historical period, for a total of 12 measurements, yielding the following initial remaining capacity values ​​(%): {99.5%, 99.3%, 99.4%, 98.8%, 99.2%, 99.1%, 99.0%, 98.9%, 98.7%, 98.6%, 98.8%, 99.0%}. The average of these data is then calculated as: (99.5+99.3+99.4+98.8+99.2+99.1+99.0+98.9+98.7+98.6+98.8+99.0) / 12≈99.0%. This average value is the average initial remaining capacity of the battery during the first historical period and can serve as an important benchmark for subsequent assessments of battery health.

[0089] This embodiment provides data support for long-term monitoring and analysis of battery health status by calculating the initial remaining capacity.

[0090] To accurately determine the operating status of the target battery, optionally, in the battery status prediction method provided in this application embodiment, determining the operating status of the target battery based on the initial remaining capacity value includes: obtaining the nominal capacity of the target battery, dividing the initial remaining capacity value by the nominal capacity to obtain the health status index value of the target battery; determining whether the health status index value is less than the index threshold; if the health status index value is less than the index threshold, determining that the operating status of the target battery is abnormal; if the health status index value is greater than or equal to the index threshold, determining that the operating status of the target battery is normal.

[0091] It should be noted that the nominal capacity, also known as the maximum amount of electricity a battery can provide under ideal conditions, is usually specified by the manufacturer and is measured in ampere-hours (Ah) or milliampere-hours (mAh). The state-of-health index, which is a percentage value obtained by dividing the battery's initial remaining capacity by its nominal capacity, reflects the battery's current health and performance.

[0092] Specifically, when determining the battery's operating status, the first step is to accurately obtain the target battery's nominal capacity. Then, the average of the initial remaining capacity is divided by the nominal capacity to calculate the health status index. For example, if the nominal capacity is 1000mAh and the average of the initial remaining capacity is 980mAh, the health status index is (980 / 1000 = 0.98) or (98%).

[0093] Next, the calculated health status index values ​​can be compared with preset threshold values. If the health status index value is lower than the threshold, it indicates that the battery performance has declined to an abnormal level, and the battery operating status should be marked as abnormal; conversely, if the health status index value is not lower than the threshold, it indicates that the battery's health status is acceptable, and its operating status is judged as normal. Through this judgment process, abnormal battery performance can be identified in a timely manner, providing necessary early warning signals for the data center's power supply system and avoiding power outages caused by battery failure.

[0094] This embodiment achieves accurate assessment of the target battery's operating status by quantitatively analyzing the battery's health status indicators and comparing them with preset indicator thresholds, thereby ensuring the normal operation of the data center.

[0095] Optionally, in the battery state prediction method provided in the embodiments of this application, the method further includes: determining the cause of the target battery abnormality when the target battery is determined to be abnormal; generating alarm information and maintenance information for maintaining the target battery based on the cause of the abnormality, and sending the alarm information and maintenance information to the operation and maintenance terminal that is related to the cause of the abnormality.

[0096] It should be noted that the causes of abnormality, that is, the specific factors that lead to abnormal battery operation, may include, but are not limited to, battery aging, internal short circuit, charging system failure, excessively high or low temperature, abnormal battery voltage, etc.

[0097] Specifically, once a target battery is identified as abnormal, further data analysis or sensor monitoring is needed to diagnose the specific cause of the abnormality. This can include checking the battery's operating data such as temperature, voltage, and charging / discharging current, comparing these data with historical normal ranges, or using battery diagnostic equipment for more in-depth testing. Accurately identifying the cause of the abnormality helps to take targeted maintenance measures, avoid additional damage that may be caused by blind operation, prevent similar failures from recurring, and improve the stability of the data center's power system.

[0098] Furthermore, after determining the cause of the anomaly, detailed alarm information and maintenance recommendations can be generated. The alarm information should include the battery's identification information, the specific manifestations of the anomaly, possible causes, and the level of urgency. The maintenance information should provide specific maintenance steps, recommended spare parts or tools, and maintenance priorities, ensuring that the maintenance team can quickly understand the battery's abnormal condition and take appropriate emergency responses. It should also guide the maintenance team to perform professional and effective repair or replacement operations to restore the battery to its normal operating state.

[0099] This embodiment accurately identifies the cause of anomalies when a target battery is found to be abnormal, generates targeted alarms and maintenance information, and quickly notifies the operations and maintenance department responsible for handling the cause of the anomaly. This effectively improves the efficiency and accuracy of battery fault handling, prevents further deterioration of battery faults, and enhances the operational security and efficiency of the entire data center.

[0100] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0101] Example 2

[0102] This application also provides a battery state prediction device. It should be noted that the battery state prediction device of this application can be used to execute the battery state prediction method provided in the above embodiments. The following describes the battery state prediction device provided in this application.

[0103] According to an embodiment of this application, an apparatus for implementing the above-described battery state prediction method is also provided. Figure 4 This is a schematic diagram of a battery state prediction device according to Embodiment 2 of this application, as shown below. Figure 4 As shown, the device includes:

[0104] The first acquisition unit 41 is used to acquire the operating data of the target battery at different historical moments within the first historical period, and obtain N sets of first historical operating data. Each set of first historical operating data includes M operating data of the target battery at the corresponding historical moment, where M is the number of types of indicator data of the target battery, and M and N are positive integers.

[0105] The prediction unit 42 is used to input N sets of first historical operating data into the operating data prediction model to obtain the operating data set of the target battery at H target times within the target time period, and obtain H operating data sets, where H is a positive integer.

[0106] The first generation unit 43 is used to generate M operating data sequences of the target battery within the target time period based on H sets of operating data.

[0107] The first determining unit 44 is used to determine the target battery as abnormal when there is an abnormal fluctuation index among the fluctuation indicators of M running data sequences.

[0108] The second determining unit 45 is used to obtain the initial remaining capacity value of the target battery in the first historical period when there is no abnormal fluctuation index among the fluctuation indexes of the M operating data sequences, and to determine the operating status of the target battery based on the initial remaining capacity value.

[0109] The battery state prediction device provided in this application embodiment acquires the operating data of the target battery at different historical moments within a first historical period by a first acquisition unit 41, obtaining N sets of first historical operating data. Each set of first historical operating data includes M operating data of the target battery at the corresponding historical moment, where M is the number of types of indicator data of the target battery, and M and N are positive integers. The prediction unit 42 inputs the N sets of first historical operating data into the operating data prediction model to obtain a set of operating data of the target battery at H target moments within a target period, resulting in H sets of operating data, where H is a positive integer. The first generation unit 43 generates M operating data sequences of the target battery within the target period based on the H sets of operating data. If there are abnormal fluctuation indicators among the fluctuation indicators of the M sets of operating data, the first determination unit 44 determines that the target battery is abnormal. If there are no abnormal fluctuation indicators among the fluctuation indicators of the M sets of operating data, the second determination unit 45 acquires the initial remaining capacity value of the target battery within the first historical period and determines the operating state of the target battery based on the initial remaining capacity value. By using a deep learning prediction model to predict operational data for a target period, and then determining the operational data sequence of various indicators based on the operational data, and by analyzing the abnormal fluctuations of the operational data sequence, the goal of accurately identifying battery status is achieved. This enables the optimization of battery replacement strategies, reduces resource waste, and lowers costs, thereby solving the technical problem of low accuracy in determining battery status based on usage time in related technologies.

[0110] Optionally, in the battery state prediction device provided in this application embodiment, the operating data prediction model is trained in the following manner: a second acquisition unit is used to acquire the operating data of the target battery in P second historical time periods to obtain P sets of second historical operating data, where P is a positive integer, and each set of second historical operating data includes Q sets of second historical operating data in the second historical time period, where Q is a positive integer; a third acquisition unit is used to acquire the third historical time period adjacent to each second historical time period, where each third historical time period is later than the adjacent second historical time period, and acquire the operating data in each third historical time period to obtain P sets of third historical operating data; a combination unit is used to combine the second historical operating data set and the third historical operating data set with adjacent relationships into a set of sample data to obtain P sets of first sample data; and a training unit is used to train the initial prediction model using the P sets of first sample data to obtain the operating data prediction model, wherein the initial prediction model is a neural network model with multiple output layers.

[0111] Optionally, in the battery state prediction device provided in the embodiments of this application, the device further includes: a calculation unit, used to calculate the dispersion of any running data sequence and obtain a dispersion threshold of the running data sequence, wherein the dispersion is at least one of the following: variance, standard deviation; a comparison unit, used to compare the dispersion with the dispersion threshold, and determine the fluctuation index of the running data sequence as an abnormal fluctuation index when the dispersion is greater than the dispersion threshold; and a third determination unit, used to determine the abnormal identification result of the running data sequence based on the value of the running data in the running data sequence when the dispersion is less than or equal to the dispersion threshold.

[0112] Optionally, in the battery state prediction device provided in this application embodiment, the third determining unit includes: a first judging module, used to obtain the data threshold range of the running data sequence and determine whether there is abnormal data outside the data threshold range in the running data sequence; a first determining module, used to determine the fluctuation index of the running data sequence as an abnormal fluctuation index when there is abnormal data; and a second determining module, used to determine the fluctuation index of the running data sequence as a normal fluctuation index when there is no abnormal data.

[0113] Optionally, in the battery state prediction device provided in the embodiments of this application, the second determining unit 45 includes: a first acquisition module, used to acquire the initial remaining capacity value of the target battery at each historical moment in the first historical period, to obtain N initial remaining capacity values; and a calculation module, used to calculate the average value of the N initial remaining capacity values ​​to obtain the initial remaining capacity value.

[0114] Optionally, in the battery state prediction device provided in this application embodiment, the second determining unit 45 includes: a second acquisition module, used to acquire the nominal capacity of the target battery and divide the initial remaining capacity value by the nominal capacity to obtain the health state index value of the target battery; a second judgment module, used to determine whether the health state index value is less than the index threshold; a third determining module, used to determine that the operating state of the target battery is abnormal when the health state index value is less than the index threshold; and a fourth determining module, used to determine that the operating state of the target battery is normal when the health state index value is greater than or equal to the index threshold.

[0115] Optionally, in the battery state prediction device provided in the embodiments of this application, the device further includes: a fourth determining unit, used to determine the cause of the target battery abnormality when the target battery is determined to be abnormal; and a second generating unit, used to generate alarm information and maintenance information for maintaining the target battery according to the cause of the abnormality, and send the alarm information and maintenance information to the operation and maintenance terminal that is related to the cause of the abnormality.

[0116] It should be noted that the first acquisition unit 41, prediction unit 42, first generation unit 43, first determination unit 44, and second determination unit 45 mentioned above correspond to steps S201 to S205 in Embodiment 1. The instances and application scenarios implemented by each of the above modules or units are the same as those implemented by the corresponding steps, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above modules can also be part of a device and can run in the computer terminal 10 provided in Embodiment 1.

[0117] Example 3

[0118] Embodiments of this application may provide an electronic device. Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 (Only one is shown) processor 1002, memory 1004, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0119] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0120] The processor can access information and applications stored in the memory via a transmission device to execute the following steps: Acquire operational data of the target battery at different historical moments within a first historical period, resulting in N sets of first historical operational data. Each set of first historical operational data includes M operational data points of the target battery at the corresponding historical moment, where M is the number of types of indicator data for the target battery, and M and N are positive integers. Input the N sets of first historical operational data into an operational data prediction model to obtain H sets of operational data for the target battery at target moments within a target period, resulting in H sets of operational data, where H is a positive integer. Generate M operational data sequences of the target battery within the target period based on the H sets of operational data. If abnormal fluctuation indicators exist among the fluctuation indicators of the M operational data sequences, determine that the target battery is abnormal. If no abnormal fluctuation indicators exist among the fluctuation indicators of the M operational data sequences, acquire the initial remaining capacity value of the target battery within the first historical period, and determine the operational status of the target battery based on the initial remaining capacity value.

[0121] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: acquire the operating data of the target battery in P second historical time periods to obtain P sets of second historical operating data, where P is a positive integer, and each set of second historical operating data includes Q sets of second historical operating data in the second historical time period, where Q is a positive integer; acquire the third historical time period adjacent to each second historical time period, where each third historical time period is later than the adjacent second historical time period, and acquire the operating data in each third historical time period to obtain P sets of third historical operating data; combine the second historical operating data sets and the third historical operating data sets with adjacent relationships into a set of sample data to obtain P sets of first sample data; use the P sets of first sample data to train the initial prediction model to obtain the operating data prediction model, where the initial prediction model is a neural network model with multiple output layers.

[0122] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: For any running data sequence, calculate the dispersion of the running data sequence and obtain the dispersion threshold of the running data sequence, wherein the dispersion is at least one of the following: variance, standard deviation; compare the dispersion with the dispersion threshold, and if the dispersion is greater than the dispersion threshold, determine the fluctuation index of the running data sequence as an abnormal fluctuation index; if the dispersion is less than or equal to the dispersion threshold, determine the anomaly identification result of the running data sequence based on the value of the running data in the running data sequence.

[0123] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain the data threshold range of the running data sequence, and determine whether there is abnormal data outside the data threshold range in the running data sequence; if there is abnormal data, determine the fluctuation index of the running data sequence as an abnormal fluctuation index; if there is no abnormal data, determine the fluctuation index of the running data sequence as a normal fluctuation index.

[0124] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: obtain the initial remaining capacity value of the target battery at each historical moment in the first historical period to obtain N initial remaining capacity values; calculate the average of the N initial remaining capacity values ​​to obtain the initial remaining capacity value.

[0125] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain the nominal capacity of the target battery, divide the initial remaining capacity value by the nominal capacity to obtain the health status index value of the target battery; determine whether the health status index value is less than the index threshold; if the health status index value is less than the index threshold, determine that the operating state of the target battery is abnormal; if the health status index value is greater than or equal to the index threshold, determine that the operating state of the target battery is normal.

[0126] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: if the target battery is found to be abnormal, determine the cause of the abnormality of the target battery; generate alarm information and maintenance information for maintaining the target battery according to the cause of the abnormality, and send the alarm information and maintenance information to the operation and maintenance terminal that is related to the cause of the abnormality.

[0127] Those skilled in the art will understand that Figure 5 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 5This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 5 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 5 The different configurations shown.

[0128] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0129] Example 4

[0130] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the battery state prediction method provided in Embodiment 1.

[0131] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0132] Embodiments of this application also provide a computer program product, which, when executed on a data processing device, is adapted to perform the steps of a method for predicting battery state.

[0133] Embodiments of this application also provide a computer-readable storage medium, which includes a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the above-described battery state prediction method.

[0134] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0135] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0136] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0137] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0138] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0139] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0140] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for predicting battery state, characterized in that, include: Obtain the operating data of the target battery at different historical moments within the first historical period to obtain N sets of first historical operating data. Each set of first historical operating data includes M operating data of the target battery at the corresponding historical moment, where M is the number of types of indicator data of the target battery, and M and N are positive integers. The N sets of first historical operating data are input into the operating data prediction model to obtain the operating data set of the target battery at H target times within the target time period, resulting in H operating data sets, where H is a positive integer; Based on the H sets of operational data, generate M sequences of operational data for the target battery within the target time period; If there are abnormal fluctuation indicators among the fluctuation indicators of the M operating data sequences, the target battery is determined to be abnormal; If there are no abnormal fluctuation indicators among the fluctuation indicators of the M operating data sequences, the initial remaining capacity value of the target battery in the first historical period is obtained, and the operating status of the target battery is determined based on the initial remaining capacity value.

2. The method according to claim 1, characterized in that, The operational data prediction model is trained in the following manner: The target battery is obtained in P second historical time periods to obtain P sets of second historical operation data, where P is a positive integer, and each set of second historical operation data includes Q sets of second historical operation data in the second historical time period, where Q is a positive integer; Obtain the third historical time period adjacent to each second historical time period, wherein each third historical time period is later than the adjacent second historical time period, and obtain the running data within each third historical time period to obtain P sets of third historical running data; The second and third historical operation data sets with adjacent relationships are combined into a set of sample data to obtain the first set of sample data, P; The initial prediction model is trained using the first sample data of the P group to obtain the running data prediction model, wherein the initial prediction model is a neural network model with multiple output layers.

3. The method according to claim 1, characterized in that, The method further includes: For any given running data sequence, calculate the dispersion of the running data sequence and obtain the dispersion threshold of the running data sequence, wherein the dispersion is at least one of the following: variance, standard deviation; The dispersion is compared with the dispersion threshold. If the dispersion is greater than the dispersion threshold, the fluctuation index of the running data sequence is determined to be an abnormal fluctuation index. If the dispersion is less than or equal to the dispersion threshold, the anomaly identification result of the running data sequence is determined based on the numerical value of the running data in the running data sequence.

4. The method according to claim 3, characterized in that, Determining the anomaly identification result of the running data sequence based on the numerical values ​​of the running data in the running data sequence includes: Obtain the data threshold range of the running data sequence, and determine whether there is any abnormal data in the running data sequence that is outside the data threshold range; In the presence of the aforementioned abnormal data, the fluctuation index of the running data sequence is determined to be an abnormal fluctuation index. In the absence of the aforementioned abnormal data, the fluctuation index of the running data sequence is determined to be a normal fluctuation index.

5. The method according to claim 1, characterized in that, Obtaining the initial remaining capacity value of the target battery during the first historical period includes: Obtain the initial remaining capacity value of the target battery at each historical moment within the first historical period to obtain N initial remaining capacity values; The average of the N initial remaining capacity values ​​is calculated to obtain the initial remaining capacity value.

6. The method according to claim 1, characterized in that, Determining the operating state of the target battery based on the initial remaining capacity value includes: Obtain the nominal capacity of the target battery, and divide the initial remaining capacity value by the nominal capacity to obtain the health status index value of the target battery; Determine whether the health status indicator value is less than the indicator threshold; If the health status indicator value is less than the indicator threshold, the operating status of the target battery is determined to be abnormal. If the health status indicator value is greater than or equal to the indicator threshold, the operating status of the target battery is determined to be without abnormalities.

7. The method according to claim 1, characterized in that, The method further includes: If the target battery is determined to be abnormal, the cause of the abnormality shall be determined. Based on the cause of the anomaly, an alarm message and maintenance information for maintaining the target battery are generated, and the alarm message and maintenance information are sent to the operation and maintenance terminal that is related to the cause of the anomaly.

8. A device for predicting battery state, characterized in that, include: The first acquisition unit is used to acquire the operating data of the target battery at different historical moments within the first historical period, and obtain N sets of first historical operating data. Each set of first historical operating data includes M operating data of the target battery at the corresponding historical moment, where M is the number of types of indicator data of the target battery, and M and N are positive integers. The prediction unit is used to input the N sets of first historical operating data into the operating data prediction model to obtain the operating data set of the target battery at H target times within the target time period, and obtain H operating data sets, where H is a positive integer; The first generation unit is configured to generate M operational data sequences of the target battery within the target time period based on the H operational data sets; The first determining unit is used to determine that the target battery is abnormal when there is an abnormal fluctuation index among the fluctuation indicators of the M running data sequences. The second determining unit is used to obtain the initial remaining capacity value of the target battery in the first historical period when there are no abnormal fluctuation indicators among the fluctuation indicators of the M operating data sequences, and to determine the operating status of the target battery based on the initial remaining capacity value.

9. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the battery state prediction method according to any one of claims 1 to 7.

10. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, executes the battery state prediction method according to any one of claims 1 to 7.

Images