Method, device and equipment for battery detection and storage medium

Abstract

The embodiment of the invention provides a battery detection method and device, equipment and a storage medium. The method includes obtaining a target state-of-charge sequence including a plurality of states of charge of a target battery. For a charging state in a plurality of charging states, a timing association feature corresponding to the charging state is determined based on the charging state and a historical charging state at a moment before the charging state. And generating a predicted charging state sequence corresponding to the target charging state sequence based on the time sequence correlation features. And determining a detection result about whether the target battery is abnormal or not based on the target charging state sequence and the predicted charging state sequence. Therefore, the accuracy of a battery detection result can be improved. Furthermore, the utilization efficiency of the battery is improved.

Description

Technical Field

[0001] The exemplary embodiments disclosed herein generally relate to the field of computers, and more particularly to methods, apparatus, devices, and computer-readable storage media for battery detection. Background Technology

[0002] Power batteries are widely used in various fields to provide energy for equipment. For example, in industrial power systems, power batteries are used in power transmission and transformation stations to provide closing current for power units, as well as for backup power for public utilities and communication. In the electric vehicle and electric bicycle industry, power batteries replace gasoline and diesel as the power source for electric vehicles or electric bicycles. During the application of power batteries, due to factors such as external environment, manufacturing processes, and usage habits, abnormalities are inevitable. If these abnormalities are not detected in time, it may lead to problems with the efficient use of the power battery. Summary of the Invention

[0003] In a first aspect of this disclosure, a battery detection method is provided. The method includes: acquiring a target charging state sequence comprising multiple charging states of a target battery, wherein the multiple charging states correspond to multiple moments in at least a portion of the charging process of the target battery and indicate battery state information at the corresponding moments; for each charging state among the multiple charging states, determining a temporal correlation feature corresponding to that charging state based on that charging state and historical charging states at moments prior to that charging state; generating a predicted charging state sequence corresponding to the target charging state sequence based on the temporal correlation feature; and determining a detection result regarding whether an anomaly exists in the target battery based on the target charging state sequence and the predicted charging state sequence.

[0004] In a second aspect of this disclosure, an apparatus for battery detection is provided. The apparatus includes: an acquisition module configured to acquire a target charging state sequence comprising multiple charging states of a target battery, the multiple charging states corresponding to multiple moments in the charging process of the target battery and indicating battery state information at the corresponding moments; a first determination module configured to, for a charging state among the multiple charging states, determine a temporal correlation feature corresponding to the charging state based on the charging state and historical charging states at moments prior to the charging state; a generation module configured to generate a predicted charging state sequence corresponding to the target charging state sequence based on the temporal correlation feature; and a second determination module configured to determine a detection result regarding whether an anomaly exists in the target battery based on the target charging state sequence and the predicted charging state sequence.

[0005] In a third aspect of this disclosure, an electronic device is provided. The device includes at least one processing unit; and at least one memory coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit. When executed by the at least one processing unit, the instructions cause the device to perform the method of the first aspect.

[0006] In a fourth aspect of this disclosure, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program that can be executed by a processor to implement the method of the first aspect.

[0007] It should be understood that the content described in this content section is not intended to limit the key or essential features of the embodiments of this disclosure, nor is it intended to restrict the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0008] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0009] Figure 1 A schematic diagram of an example environment in which embodiments of the present disclosure can be implemented is shown;

[0010] Figure 2 An architectural diagram of an example of a battery detection system according to some embodiments of the present disclosure is shown;

[0011] Figure 3 An architecture diagram of an example training system for a battery detection system according to some embodiments of the present disclosure is shown;

[0012] Figure 4 An architectural diagram of an example of a battery detection model according to some embodiments of the present disclosure is shown;

[0013] Figure 5 A flowchart of a battery detection process according to some embodiments of the present disclosure is shown;

[0014] Figure 6 A block diagram of an apparatus for battery detection according to some embodiments of the present disclosure is shown; and

[0015] Figure 7 A block diagram of an apparatus capable of implementing several embodiments of the present disclosure is shown. Detailed Implementation

[0016] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0017] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0018] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0019] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0020] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0021] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0022] It should be noted that the headings of any section / subsection provided herein are not limiting. Various embodiments are described throughout this document, and embodiments of any type may be included under any section / subsection. Furthermore, embodiments described in any section / subsection may be combined in any way with any other embodiments described in the same section / subsection and / or different sections / subsections.

[0023] In this document, unless explicitly stated otherwise, performing a step in response to A does not mean that the step is performed immediately after A, but may include one or more intermediate steps.

[0024] In the description of embodiments of this disclosure, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The term "some embodiments" should be understood as "at least some embodiments". Other explicit and implicit definitions may also be included below. The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0025] As used in this paper, the term "model" refers to a system that learns the relationship between inputs and outputs from training data, enabling it to generate corresponding outputs for a given input after training. Model generation can be based on machine learning techniques. Deep learning is a machine learning algorithm that uses multiple layers of processing units to process inputs and provide corresponding outputs. In this paper, "model" may also be referred to as a "machine learning model," a "machine learning network," or simply a "network," and these terms are used interchangeably. A model can also include different types of processing units or networks.

[0026] As used herein, a “unit,” “operation unit,” or “subunit” can consist of any suitable machine learning model or network. As used herein, a set of elements or similar expressions can include one or more such elements. For example, “a set of convolutional units” can include one or more convolutional units.

[0027] Figure 1 A schematic diagram of an example environment 100 in which embodiments of the present disclosure can be implemented is shown. For example... Figure 1 As shown, environment 100 includes a power battery 120, a charging device, and electronic equipment 110. A battery management system 130 is deployed in electronic equipment 110. The power battery 120 includes multiple individual cells (also referred to as battery cells).

[0028] The charging device is used to charge the power battery 120 and can record charging data of the power battery 120 during the charging process, such as the voltage, temperature, capacity, and state of charge (SOC) of each individual cell during charging. The charging device can include any known or future available type, such as a charging pile or charging cabinet, and the embodiments of this disclosure are not limited thereto. The charging device can communicate with the battery management system 130 to send charging data to the battery management system 130. The battery management system 130 can determine the operating status of the power battery 120 based on the received charging data. The battery management system 130 can include any known or future available type, such as a BMS battery system or a cloud platform, and the embodiments of this disclosure are not limited thereto. In the embodiments of this disclosure, the battery management system 130 is mainly used for charging based on data (e.g., voltage, temperature, and current). It should be understood that the embodiments of this disclosure can also be applied to other types of batteries, such as lithium manganese oxide batteries.

[0029] In environment 100, electronic device 110 can be any type of computing device, including terminal devices or server devices. Terminal devices can be any type of mobile terminal, fixed terminal, or portable terminal, including mobile phones, desktop computers, laptop computers, notebook computers, netbook computers, tablet computers, media computers, multimedia tablets, personal communication system (PCS) devices, personal navigation devices, personal digital assistants (PDAs), audio / video players, digital cameras / camcorders, positioning devices, television receivers, radio receivers, e-book devices, gaming devices, or any combination of the foregoing, including accessories and peripherals of these devices or any combination thereof. Server devices may include, for example, computing systems / servers, such as mainframes, edge computing nodes, electronic device 110 in a cloud environment, etc.

[0030] It should be understood that the structure and function of environment 100 are described for illustrative purposes only and do not imply any limitation on the scope of this disclosure.

[0031] The following description will continue with reference to the accompanying drawings, which will provide some exemplary embodiments of this disclosure.

[0032] As briefly mentioned earlier, power batteries are susceptible to abnormal risks during application. To ensure timely detection of these abnormalities and the implementation of maintenance measures, thereby improving battery utilization efficiency, state monitoring of power batteries is crucial. Due to the complexity of battery mechanisms, current methods primarily employ data-driven and deep model-based approaches to detect the state of power batteries.

[0033] A novel unsupervised method for detecting anomalies in power batteries (DyAD) has been proposed, treating the power battery as a dynamic system. DyAD uses information such as the total current output of each individual cell as system input and the total voltage, highest and lowest individual cell temperatures, highest and lowest individual cell voltages as system responses to determine if the battery is in an abnormal state. DyAD employs a Variational Autoencoder (VAE) to perform encoding and decoding operations on the system input and response. Ultimately, the abnormal state of the power battery is assessed based on the reconstruction error. However, DyAD's power battery anomaly detection technology is significantly affected by the final state of the charging process, easily overlooking information from the earlier stages of charging, leading to inaccurate detection of battery anomalies.

[0034] Embodiments of this disclosure propose a scheme for battery detection. According to various embodiments of this disclosure, a target charging state sequence comprising multiple charging states of a target battery is acquired, each charging state corresponding to multiple moments in at least a portion of the charging process of the target battery and indicating battery state information at the corresponding moment. For each charging state, a temporal correlation feature corresponding to that charging state is determined based on that charging state and historical charging states at moments prior to that charging state. Based on the temporal correlation feature, a predicted charging state sequence corresponding to the target charging state sequence is generated. Based on the target charging state sequence and the predicted charging state sequence, a detection result regarding the presence of an anomaly in the target battery is determined.

[0035] In this way, for each charging state in the target charging state sequence, corresponding temporal correlation features are determined based on the charging state and the historical charging states at previous times. A predicted charging sequence is generated based on these temporal correlation features to determine the detection result. Therefore, determining whether a battery is abnormal based on the correlation features between charging states at different times improves the accuracy of battery detection results compared to determining the detection result solely based on the last time state. Furthermore, this improves battery utilization efficiency.

[0036] Figure 2 An architectural diagram of an example of a battery detection system 200 according to some embodiments of the present disclosure is shown. Figure 2 In the example, the battery detection system 200 can be implemented or included in the electronic device 110.

[0037] In some embodiments, the electronic device 110 acquires a target charging state sequence 210 including multiple charging states of a target battery. The multiple charging states correspond to multiple moments in at least a portion of the charging process of the target battery and indicate battery state information at the corresponding moments. For example, the charging state at a certain moment indicates an encoded representation of the charging data acquired at that moment, such as a low-dimensional representation. The charging data includes state information of multiple individual cells in the target battery. The charging data may include two types of information: system response 211 and system input 212.

[0038] In some embodiments, data can be collected during the charging process of the target battery. Following national standard protocols, charging data such as timestamps, battery state of charge, total current, total voltage, highest single-cell temperature, lowest single-cell temperature, highest single-cell voltage, and lowest single-cell voltage are collected at a fixed frequency (e.g., every 10 seconds) during the charging process of the power battery (120). In some embodiments, the collected charging data can also be uploaded to a cloud server and stored to facilitate subsequent detection of battery anomalies based on the battery's charging data. Simultaneously, static information of the battery (e.g., battery type, rated voltage, rated capacity, production date, etc.) is retained to determine the type of target battery.

[0039] In some embodiments, anomalies may occur during the acquisition, transmission, and parsing phases of the collected information. To improve the accuracy and reliability of the detection result 270, preprocessing of the collected battery state is necessary. For example, preprocessing the battery state includes setting data exceeding specified boundaries to null values, based on the reasonable range of various types of data during the operation of the power battery 120 as defined by national standards (e.g., SOC between 1 and 100). The electronic device 110 can filter the collected battery state, sort the charging data by timestamp, and remove duplicate data. Furthermore, if missing data exists due to acquisition gaps or nulling operations exceeding specified boundaries, the mean of preceding and subsequent data can be used to fill the missing data.

[0040] In some embodiments, to accelerate the convergence speed of the neural network and eliminate dimensional differences between features, it is necessary to perform standardization processing on the charging states in the target charging state sequence 210. The formulas for performing standardization operations on fields such as total voltage and total current are as follows:

[0041]

[0042] Where z represents the standardized data, x represents the original data, μ represents the mean of the data, and σ represents the standard deviation.

[0043] The initial charging state sequence corresponds to all charging states during a single charging process of the target battery. When the charging process has a long time span, the model focuses more on the average error of reconstructing the entire charging sequence, resulting in lower attention to local charging sequences, thus affecting the accuracy and reliability of the detection results for the target battery. Therefore, in some embodiments, the electronic device 110 can divide the initial charging state sequence into multiple short first charging state sequences. In some embodiments, the electronic device 110 acquires an initial charging state sequence related to the target battery, where the charging states in the initial charging state sequence correspond to multiple moments during the charging process of the target battery. The initial charging state sequence is divided into one or more first charging state sequences according to a predetermined time window. And based on one or more first charging state sequences, a target charging state sequence 210 is determined.

[0044] For example, the electronic device 110 extracts a short sequence of predetermined length starting from the charging start position using a time window of predetermined length. Then, it moves backward from the start position according to a predetermined step size, repeating the operation until the last first charging state sequence reaches the position where the initial charging sequence ends. Thus, multiple first charging state sequences related to the initial charging state sequence are obtained. At least one charging state sequence among the multiple first charging state sequences is used as the target charging state sequence 210. Repeated charging states may exist between the various first charging state sequences.

[0045] In some embodiments, the state of the target battery is determined using multiple first charging state sequences, enabling the battery detection system to focus on local information. However, this approach loses global information about the charging state, thus requiring a sampling operation. In some embodiments, sampling is performed on the initial charging sequence to determine a second charging state sequence of the same predetermined length. The charging state sequence from one or more first and second charging state sequences is determined as the target charging state sequence 210.

[0046] In some embodiments, for a charging state among multiple charging states, a time-series association feature corresponding to the charging state is determined based on the charging state and historical charging states at times prior to the charging state.

[0047] The battery detection model 280 can be implemented based on a gated recurrent unit (GRU). The battery detection model 280 performs encoding and decoding operations on the target charging state sequence 210 using an encoder 220 and a decoder 240 implemented based on the GRU. During the decoding phase, the battery detection model 280 performs the decoding operation only based on the hidden state in the last time step.

[0048] like Figure 2As shown, the battery detection model 280 acquires the provided target charging state sequence 210 and performs an encoding operation on the target charging state sequence 210 using the encoder 220. The battery detection model 280 determines the temporal association features corresponding to each charging state through the attention module 230. The temporal association features indicate the contextual features of the charging state. In some embodiments, the electronic device 110 performs an encoding operation on the target charging state sequence 210 using the encoder 220 in the battery detection model 280 to obtain multiple encoded representations corresponding to multiple charging states (i.e., the hidden state of the encoder 220 at each time step). Subsequently, the electronic device 110 determines the temporal association features corresponding to the charging state based on the target encoded representation and the weights of the charging state and historical charging states.

[0049] Figure 3 An architectural diagram of an example of a battery detection model 280 according to some embodiments of the present disclosure is shown. Figure 3 As shown, the target charging state sequence 210 includes charging states X = (x1, x2, ..., x...). T At time step t, encoder 220 inputs x. t Perform encoding operations to obtain the corresponding encoded representation (h1, h2, ..., h T Subsequently, the encoded representation corresponding to each charging state in the target charging state sequence 210 is input into the attention module 230 to obtain the corresponding attention vector 410. Encoded representation h i corresponding attention vector The calculation formula is as follows:

[0050]

[0051] in W d U d Here are the model parameters, and tanh is the activation function. This represents the concatenation result of the hidden state and cell state of decoder 240 at the previous time step, where T represents the total time step length, and h represents the total time step length. i This represents the encoded representation of encoder 220 at time step i.

[0052] Subsequently, the electronic device 110 uses the softmax function to determine the weights of each encoded representation. Encoded representation h i weight The calculation formula is as follows:

[0053]

[0054] Electronic device 110 determines the temporal correlation features (i.e., the context vector at time t) corresponding to the charging state based on the target encoded representation and the weights of the current charging state and historical charging states. This determines the information fusion of encoder 220 at different time steps. The temporal correlation feature c corresponding to the encoded representation at time step i is... t The calculation formula is as follows:

[0055]

[0056] continue Figure 2 In some embodiments, the electronic device 110 generates a predicted charging state sequence 250 corresponding to the target charging state sequence 210 based on temporal correlation features.

[0057] Electronic device 110 generates a predicted charging state sequence 250 using decoder 240 based on temporal association features corresponding to multiple charging states. In some embodiments, decoder 240 is a model implemented based on GRU. Generating the predicted charging state sequence 250 using decoder 240 includes multiple generation steps. Decoder 240 performs a decoding operation based on the hidden state generated in the generation step corresponding to the previous time step and the corresponding temporal association features to determine the predicted charging state and hidden state corresponding to the charging state.

[0058] In some embodiments, the electronic device 110 acquires the preceding hidden state and the preceding predicted charging state generated in the previous generation step for a given generation step. Based on the time corresponding to the given generation step, a given temporal correlation feature is determined. Based on the preceding hidden state, the preceding predicted charging state, and the given temporal correlation feature, the decoder 240 generates the predicted charging state and the hidden state in the given generation step.

[0059] like Figure 3 As shown, for x in the target charging state sequence t In the t-th time step of the battery detection model, the electronic device 110 first determines its corresponding time-series correlation feature C. t Subsequently, electronic device 110 uses a decoder to analyze the timing-related feature C. t The hidden state Ct generated by the decoder in time step t-1 and the predicted charging state output by the encoder in time step t-1 are used to generate the predicted charging state corresponding to the current time step (i.e., time step t). This iterative process is repeated to finally obtain the predicted charging state sequence Y = (y1, y2, ..., y...). T ).

[0060] In some embodiments, to improve the accuracy of model detection, the battery detection model can also be trained to update the model parameters. Figure 4An architectural diagram of an example training system for a battery detection system according to some embodiments of the present disclosure is shown. Figure 4 As shown, the electronic device 110 trains the model using the target charging sequence to generate a training predicted charging state sequence. Subsequently, based on the difference between the training predicted charging state sequence and the target charging state sequence, a loss function value including reconstruction loss and KL loss is determined. In some embodiments, mileage prediction can be introduced as weak label supervision, ultimately generating a weak label loss. The electronic device 110 determines the model training loss based on the reconstruction loss, KL loss, and weak label loss. These three losses have inconsistent importance, so weight coefficients need to be defined between them. Larger weights can be set for the reconstruction error loss and the KL loss. In some embodiments, Adam can be selected as the optimizer, and training begins with the prepared training data, using only normal charging sequences during training.

[0061] continue Figure 2 In some embodiments, the electronic device 110 determines a detection result 270 regarding whether an anomaly exists in the target battery based on a target charging state sequence 210 and a predicted charging state sequence 250.

[0062] For example, electronic device 110 first determines the reconstruction difference 260 between the target charging state sequence 210 and the predicted charging state sequence 250. Then, electronic device 110 determines whether the reconstruction difference 260 is greater than a target anomaly threshold corresponding to the battery. If it is greater than the target anomaly threshold, the detection result 270 indicates that the target battery has an anomaly.

[0063] In some embodiments, the electronic device 110 may first determine the type of target battery based on the battery parameters of the target battery. Subsequently, a target anomaly threshold is determined based on the type of target battery and the anomaly thresholds corresponding to each of the multiple battery types.

[0064] Due to the differences between different types of power batteries 120, different target anomaly thresholds can be provided for different types of power batteries 120. In some embodiments, for a battery type among multiple battery types, multiple reference state-of-charge sequences of batteries having that battery type are determined. Based on multiple reference reconstruction errors corresponding to the multiple reference state-of-charge sequences, a first number of abnormal samples and a second number of normal samples are determined. Based on the multiple reference reconstruction errors, the anomaly threshold corresponding to that battery type is determined by maximizing the ratio of the first number to the second number.

[0065] For example, static information about the battery, including battery type, production date, and service life, is numerically encoded. Then, a clustering algorithm (such as K-means) is used to classify the batteries into different categories. The reconstruction errors are grouped according to the battery category of the cluster, and then the errors are sorted from largest to smallest. The quantile that results in the highest proportion of abnormal samples compared to normal samples is selected as the anomaly threshold.

[0066] As can be seen, to address the issue of insufficient attention to local charging sequences in the input data, the input time series is segmented and sampled to obtain one or more first battery charging sequences and second battery charging sequences. This allows the model to maintain attention to both local and global information throughout the entire charging process during learning. To address the issue of focusing only on the last time state of the encoder while ignoring earlier time states during the decoding stage, an attention mechanism is introduced. This ensures that the latent vectors obtained during decoding include vectors from all historical time states. By selectively focusing on important time state information, the attention mechanism can automatically learn the importance of different time state vectors of the encoder for decoding. To address the issue of using a uniform threshold for all power batteries during the evaluation stage while ignoring differences between different power batteries, clustering is performed based on the static information of the power batteries, and then different thresholds are assigned to different battery categories. This improves the accuracy and reliability of battery state detection results.

[0067] Figure 5 A flowchart of a process 500 for battery detection according to some embodiments of the present disclosure is shown. Process 500 can be implemented at electronic device 110.

[0068] In box 510, a target charging state sequence including multiple charging states of the target battery is obtained, wherein the multiple charging states correspond to multiple moments of at least a portion of the charging process of the target battery and indicate the battery state information at the corresponding moment.

[0069] In some embodiments, obtaining a target charging state sequence including multiple charging states of a target battery includes: obtaining an initial charging state sequence related to the target battery, wherein the charging states in the initial charging state sequence correspond to multiple moments in the charging process of the target battery; dividing the initial charging state sequence into one or more first charging state sequences according to a time window of a predetermined length; and determining a target charging state sequence based on one or more first charging state sequences.

[0070] In some embodiments, determining the target charging state sequence includes: sampling an initial charging sequence to determine a second charging state sequence of the same length; and determining a charging state sequence from one or more first charging state sequences and second charging state sequences as the target charging state sequence.

[0071] In box 520, for a charging state among multiple charging states, based on the charging state and the historical charging states at times before the charging state, the temporal association features corresponding to the charging state are determined.

[0072] In some embodiments, determining the temporal association features corresponding to the charging state includes: performing an encoding operation on the target charging state sequence using an encoder to obtain multiple encoded representations corresponding to multiple charging states respectively; and determining the temporal association features corresponding to the charging state based on the target encoded representations and the weights of the charging state and historical charging states.

[0073] In box 530, a predicted charging state sequence corresponding to the target charging state sequence is generated based on temporal correlation features.

[0074] In some embodiments, generating a predicted charging state sequence corresponding to the target charging state sequence based on temporal correlation features includes: generating a predicted charging state sequence using a decoder based on temporal correlation features corresponding to multiple charging states respectively.

[0075] In some embodiments, generating a predicted charging state sequence includes multiple generation steps, and a given generation step among the multiple generation steps includes: obtaining a preceding hidden state and a preceding predicted charging state generated by the previous generation step; determining a given temporal correlation feature based on the time corresponding to the given generation step; and generating the predicted charging state and hidden state in the given generation step using a decoder based on the preceding hidden state, the preceding predicted charging state, and the given temporal correlation feature.

[0076] In box 540, based on the target state of charge sequence and the predicted state of charge sequence, a detection result is determined regarding whether there is an anomaly in the target battery.

[0077] In some embodiments, determining the detection result regarding whether the target battery is abnormal includes: determining the reconstruction difference between the target charging state sequence and the predicted charging state sequence; and determining that the detection result indicates that the target battery is abnormal in response to the reconstruction difference exceeding a target abnormality threshold.

[0078] In some embodiments, the target anomaly threshold is determined by: determining the type of the target battery based on the battery parameters of the target battery; and determining the target anomaly threshold based on the type of the target battery and the anomaly thresholds corresponding to the multiple battery types.

[0079] In some embodiments, the anomaly thresholds corresponding to the multiple battery types are determined by: for each battery type among the multiple battery types, determining multiple reference charging state sequences of batteries having that battery type; determining a first number of abnormal samples and a second number of normal samples based on multiple reference reconstruction errors corresponding to the multiple reference charging state sequences; and determining the anomaly threshold corresponding to that battery type by maximizing the ratio of the first number to the second number based on the multiple reference reconstruction errors.

[0080] Figure 6 A schematic structural block diagram of a battery detection device 600 according to certain embodiments of the present disclosure is shown. The device 600 may be implemented as or included in an electronic device 110. Various modules / components in the device 600 may be implemented by hardware, software, firmware, or any combination thereof.

[0081] As shown in the figure, the device 600 includes an acquisition module 610, configured to acquire a target charging state sequence including multiple charging states of a target battery. Each charging state corresponds to a different moment in the charging process of the target battery and indicates the battery state information at that moment. The device 600 also includes a first determination module 620, configured to determine a temporal correlation feature corresponding to a charging state based on that charging state and historical charging states prior to that charging state. The device 600 further includes a generation module 630, configured to generate a predicted charging state sequence corresponding to the target charging state sequence based on the temporal correlation feature. The device 600 also includes a second determination module 640, configured to determine a detection result regarding the presence of an anomaly in the target battery based on the target charging state sequence and the predicted charging state sequence.

[0082] In some embodiments, the acquisition module 610 is further configured to acquire an initial charging state sequence related to the target battery, wherein the charging states in the initial charging state sequence correspond to multiple moments in the charging process of the target battery; divide the initial charging state sequence into one or more first charging state sequences according to a time window of a predetermined length; and determine a target charging state sequence based on one or more first charging state sequences.

[0083] In some embodiments, the acquisition module 610 is further configured to perform sampling on the initial charging sequence to determine a second charging state sequence of the same length as the predetermined sequence; and to determine the charging state sequence in one or more of the first and second charging state sequences as the target charging state sequence.

[0084] In some embodiments, the first determining module 620 is further configured to perform an encoding operation on the target charging state sequence using an encoder to obtain multiple encoded representations corresponding to multiple charging states respectively; and to determine the temporal association features corresponding to the charging state based on the target encoded representation and the weights of the charging state and historical charging states.

[0085] In some embodiments, the generation module 630 is further configured to generate a predicted charging state sequence using a decoder based on time-series association features corresponding to multiple charging states respectively.

[0086] In some embodiments, the generation module 630 is further configured to obtain the preceding hidden state and the preceding predicted charging state generated in the previous generation step for a given generation step; determine the given temporal correlation feature based on the time corresponding to the given generation step; and generate the predicted charging state and the hidden state in the given generation step using a decoder based on the preceding hidden state, the preceding predicted charging state and the given temporal correlation feature.

[0087] In some embodiments, the second determining module 640 is further configured to determine the reconstruction difference between the target charging state sequence and the predicted charging state sequence; and in response to the reconstruction difference exceeding a target anomaly threshold, determine that the detection result indicates an anomaly in the target battery.

[0088] In some embodiments, the second determining module 640 is further configured to determine the type of target battery based on the battery parameters of the target battery; and to determine a target anomaly threshold based on the type of target battery and anomaly thresholds corresponding to multiple battery types respectively.

[0089] In some embodiments, the second determining module 640 is further configured to, for a battery type among a plurality of battery types, determine a plurality of reference charging state sequences of batteries having that battery type; determine a first number of abnormal samples and a second number of normal samples based on a plurality of reference reconstruction errors corresponding to the plurality of reference charging state sequences; and determine an abnormal threshold corresponding to that battery type by maximizing the ratio of the first number to the second number based on the plurality of reference reconstruction errors.

[0090] Figure 7 A block diagram is shown illustrating an electronic device 700 in which one or more embodiments of the present disclosure may be implemented. It should be understood that... Figure 7 The electronic device 700 shown is merely exemplary and should not be construed as limiting the functionality and scope of the embodiments described herein. Figure 7 The electronic device 700 shown can be used to achieve Figure 1 Electronic devices 110.

[0091] like Figure 7As shown, electronic device 700 is in the form of general-purpose electronic device 110. Components of electronic device 700 may include, but are not limited to, one or more processors or processing units 710, memory 720, storage device 730, one or more communication units 740, one or more input devices 750, and one or more output devices 760. Processing unit 710 may be a physical or virtual processor and is capable of performing various processes according to programs stored in memory 720. In a multiprocessor system, multiple processing units execute computer-executable instructions in parallel to improve the parallel processing capability of electronic device 700.

[0092] Electronic device 700 typically includes multiple computer storage media. Such media can be any accessible media that is accessible to electronic device 700, including but not limited to volatile and non-volatile media, removable and non-removable media. Memory 720 can be volatile memory (e.g., registers, cache, random access memory (RAM)), non-volatile memory (e.g., read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory), or some combination thereof. Storage device 730 can be removable or non-removable media and can include machine-readable media, such as flash drives, disks, or any other media that can be used to store information and / or data and can be accessed within electronic device 700.

[0093] Electronic device 700 may further include additional removable / non-removable, volatile / non-volatile storage media. Although not explicitly stated... Figure 7 As shown, disk drives for reading from or writing to removable, non-volatile disks (e.g., "floppy disks") and optical disk drives for reading from or writing to removable, non-volatile optical disks can be provided. In these cases, each drive can be connected to a bus (not shown) via one or more data media interfaces. Memory 720 may include computer program product 725 having one or more program modules configured to perform various methods or actions of various embodiments of this disclosure.

[0094] The communication unit 740 enables communication with other electronic devices 110 via a communication medium. Additionally, the functionality of the components of electronic device 700 can be implemented using a single computing cluster or multiple computing machines capable of communicating via communication connections. Therefore, electronic device 700 can operate in a networked environment using logical connections to one or more other servers, network personal computers (PCs), or another network node.

[0095] Input device 750 can be one or more input devices, such as a mouse, keyboard, trackball, etc. Output device 760 can be one or more output devices, such as a monitor, speaker, printer, etc. Electronic device 700 can also communicate with one or more external devices (not shown) via communication unit 740 as needed. External devices include storage devices, display devices, etc., and can communicate with one or more devices that enable user interaction with electronic device 700, or with any device (e.g., network card, modem, etc.) that enables electronic device 700 to communicate with one or more other electronic devices 110. Such communication can be performed via input / output (I / O) interface (not shown).

[0096] According to an exemplary implementation of this disclosure, a computer-readable storage medium is provided that stores computer-executable instructions thereon, wherein the computer-executable instructions are executed by a processor to implement the methods described above. According to an exemplary implementation of this disclosure, a computer program product is also provided, which is tangibly stored on a non-transitory computer-readable medium and includes computer-executable instructions, which are executed by a processor to implement the methods described above.

[0097] Various aspects of this disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, devices, and computer program products implemented according to this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0098] These computer-readable program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processing unit of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner. Thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0099] Computer-readable program instructions can be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions that execute on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0100] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction, which contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0101] Various implementations of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed implementations. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described implementations. The terminology used herein is chosen to best explain the principles, practical applications, or improvements to technology in the market, or to enable others skilled in the art to understand the various implementations disclosed herein.

Claims

1. A battery testing method, characterized in that, include: A target charging state sequence is obtained, which includes multiple charging states of a target battery, wherein the multiple charging states correspond to multiple moments of at least a portion of the charging process of the target battery and indicate the battery state information at the corresponding moments. For each of the multiple charging states, a time-series correlation feature corresponding to the charging state is determined based on the charging state and the historical charging states at times prior to the charging state. Based on the time-series correlation features, a predicted charging state sequence corresponding to the target charging state sequence is generated; as well as Based on the target charging state sequence and the predicted charging state sequence, a detection result is determined regarding whether the target battery has any abnormalities.

2. The battery testing method according to claim 1, characterized in that, Determining the timing association features corresponding to the charging state includes: An encoder is used to perform an encoding operation on the target charging state sequence to obtain multiple encoded representations corresponding to the multiple charging states; and Based on the target encoding representation and the weights of the charging state and the historical charging states, the temporal correlation feature corresponding to the charging state is determined.

3. The battery testing method according to claim 1, characterized in that, The step of generating a predicted charging state sequence corresponding to the target charging state sequence based on the temporal correlation features includes: Based on the temporal correlation features corresponding to the multiple charging states respectively, the decoder is used to generate the predicted charging state sequence.

4. The battery testing method according to claim 3, characterized in that, Generating the predicted charging state sequence includes multiple generation steps, and a given generation step among the multiple generation steps includes: Obtain the preceding hidden state and the preceding predicted charging state generated in the previous generation step of the given generation step; Based on the time corresponding to the given generation step, determine the given temporal correlation features; and Based on the preceding hidden state, the preceding predicted charging state, and the given temporal correlation features, the decoder is used to generate the predicted charging state and the hidden state in the given generation step.

5. The battery testing method according to claim 1, characterized in that, The acquisition of the target charging state sequence, which includes multiple charging states of the target battery, includes: Obtain an initial charging state sequence related to the target battery, wherein the charging states in the initial charging state sequence correspond to multiple moments in the charging process of the target battery; The initial charging state sequence is divided into one or more first charging state sequences according to a predetermined time window; and The target charging state sequence is determined based on the one or more first charging state sequences.

6. The battery testing method according to claim 5, characterized in that, Determining the target charging state sequence includes: Sampling is performed on the initial charging sequence to determine a second charging state sequence of the same length as the predetermined sequence; and The charging state sequence in one or more first charging state sequences and second charging state sequences is determined as the target charging state sequence.

7. The battery testing method according to claim 1, characterized in that, The detection results used to determine whether the target battery is abnormal include: Determine the reconstruction difference between the target charging state sequence and the predicted charging state sequence; and In response to the reconstruction difference exceeding the target anomaly threshold, the detection result is determined to indicate that the target battery is abnormal.

8. The battery testing method according to claim 7, characterized in that, The target anomaly threshold is determined in the following way: Based on the battery parameters of the target battery, the type of the target battery is determined; as well as The target anomaly threshold is determined based on the type of the target battery and the anomaly thresholds corresponding to each of the multiple battery types.

9. The battery testing method according to claim 8, characterized in that, The abnormal thresholds corresponding to the various battery types are determined in the following manner: For each of the multiple battery types, a multiple reference state-of-charge sequence of batteries of that type is determined; Based on multiple reference reconstruction errors corresponding to the multiple reference charging state sequences, a first number of abnormal samples and a second number of normal samples are determined. as well as Based on multiple reference reconstruction errors, the anomaly threshold corresponding to the battery type is determined by maximizing the ratio of the first number to the second number.

10. A battery testing device, characterized in that, include: The acquisition module is configured to acquire a target charging state sequence including multiple charging states of the target battery, wherein the multiple charging states correspond to multiple moments in the charging process of the target battery and indicate the battery state information at the corresponding moment. The first determining module is configured to determine, for a charging state among the plurality of charging states, a time-series association feature corresponding to the charging state based on the charging state and historical charging states at times prior to the charging state. The generation module is configured to generate a predicted charging state sequence corresponding to the target charging state sequence based on the temporal correlation features. as well as The second determining module is configured to determine a detection result regarding whether there is an anomaly in the target battery based on the target charging state sequence and the predicted charging state sequence.

11. An electronic device, comprising: At least one processing unit; as well as At least one memory, coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, which, when executed by the at least one processing unit, cause the electronic device to perform the method according to any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program stored thereon, the computer program being executable by a processor to implement the method according to any one of claims 1 to 9.

Images