Charging warning method, device, equipment, storage medium and program product

Abstract

The application relates to a charging early warning method and device, equipment, a storage medium and a program product. The method comprises the following steps: acquiring first charging operation data generated by a battery of a vehicle at a plurality of first charging time points in succession and predicted charging operation data of a first charging period; the first charging period is earlier than the plurality of first charging time points; performing screening processing on the predicted charging operation data based on the plurality of first charging time points, obtaining screened predicted charging operation data, adding actual operation data of the battery in the first charging period to the screened predicted charging operation data, and obtaining new predicted charging operation data; determining second charging operation data of the battery at a plurality of second charging time points in succession based on the first charging operation data and the new predicted charging operation data, and determining charging early warning information of the battery based on the second charging operation data. The method can improve the accuracy of charging early warning.

Description

Technical Field

[0001] This application relates to the field of charging technology, and in particular to a charging early warning method, device, equipment, storage medium, and program product. Background Technology

[0002] As the power source for new energy vehicles, the operating status of the power battery during the charging process is directly related to the vehicle's safety, battery life and charging efficiency. If problems such as abnormal voltage drift, sudden temperature rise, or excessive current fluctuation occur during the charging process, and no timely warning or intervention measures are taken, it is very easy to cause safety accidents such as battery bulging, thermal runaway or even fire and explosion. At the same time, it will also accelerate battery aging and degradation and reduce the battery's cycle life.

[0003] In existing technologies, most methods rely on threshold-based rules to provide early warnings about the charging process of vehicle power batteries. However, existing charging warning methods suffer from low accuracy. Summary of the Invention

[0004] Therefore, it is necessary to provide a charging early warning method, device, equipment, storage medium, and program product with high accuracy to address the above-mentioned technical problems.

[0005] Firstly, this application provides a charging warning method, including:

[0006] Acquire first charging operation data and predicted charging operation data of the vehicle's battery at multiple consecutive first charging moments; the first charging period is earlier than multiple first charging moments.

[0007] Based on multiple first charging moments, the predicted charging operation data is filtered to obtain the filtered predicted charging operation data. The actual operation data of the battery during the first charging period is then added to the filtered predicted charging operation data to obtain new predicted charging operation data.

[0008] Based on the first charging operation data and the new predicted charging operation data, the second charging operation data of the battery at multiple consecutive second charging moments is determined, and the charging warning information of the battery is determined based on the second charging operation data.

[0009] In one embodiment, determining the second charging operation data of the battery at multiple consecutive second charging moments based on the first charging operation data and the new predicted charging operation data includes: performing alignment resampling processing on the first charging operation data and the new predicted charging operation data based on a preset state of charge increment to obtain target charging operation data; and inputting the target charging operation data into a pre-trained prediction model for prediction processing to obtain the second charging operation data.

[0010] Secondly, this application also provides a charging warning device, comprising:

[0011] The acquisition module is used to acquire first charging operation data and predicted charging operation data of the vehicle's battery during multiple consecutive first charging moments; the first charging period is earlier than the multiple first charging moments.

[0012] The execution module is used to filter the predicted charging operation data based on multiple first charging moments to obtain the filtered predicted charging operation data, and add the actual operation data of the battery in the first charging period to the filtered predicted charging operation data to obtain new predicted charging operation data.

[0013] The determination module is used to determine the second charging operation data of the battery at multiple consecutive second charging moments based on the first charging operation data and the new predicted charging operation data, and to determine the charging warning information of the battery based on the second charging operation data.

[0014] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described in any of the embodiments of the first aspect above.

[0015] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the embodiments of the first aspect above.

[0016] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in any of the embodiments of the first aspect above.

[0017] The aforementioned charging warning method, device, equipment, storage medium, and program product first acquire first charging operation data and predicted charging operation data for a first charging period generated by the vehicle's battery at multiple consecutive first charging moments, where the first charging period is earlier than the multiple first charging moments. Then, the predicted charging operation data is filtered based on the multiple first charging moments to obtain filtered predicted charging operation data. The actual operation data of the battery in the first charging period is added to the filtered predicted charging operation data to obtain new predicted charging operation data. Finally, based on the first charging operation data and the new predicted charging operation data, the second charging operation data of the battery at multiple consecutive second charging moments is determined, and the charging warning information of the battery is determined based on the second charging operation data. The charging early warning method provided in this application predicts the charging operation data for multiple consecutive second charging moments by using first charging operation data generated at multiple consecutive first charging moments and predicted charging operation data for first charging periods earlier than multiple first charging moments. In other words, it uses the charging operation data generated in the current charging process and the charging operation data of historical charging processes to predict the charging operation data for the remaining stages of the current charging process. Since the predicted charging operation data is constantly updated, it takes into account both long-term charging patterns and adapts to short-term charging states, effectively improving the accuracy of the predicted charging operation data, and thus improving the accuracy of the charging early warning. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a charging warning method in one embodiment;

[0020] Figure 2 This is a flowchart illustrating a method for determining predicted charging operation data of a battery during the current charging process in one embodiment.

[0021] Figure 3 This is a flowchart illustrating a method for obtaining target charging operation data in one embodiment;

[0022] Figure 4 This is a flowchart illustrating a method for obtaining predicted charging operation data output by a prediction model in one embodiment.

[0023] Figure 5 This is a flowchart illustrating a method for obtaining global charging feature information in one embodiment;

[0024] Figure 6 This is a flowchart illustrating a method for obtaining predicted charging operation data in one embodiment;

[0025] Figure 7 This is a flowchart illustrating the charging warning method in another embodiment;

[0026] Figure 8 This is a structural block diagram of a charging warning device in one embodiment;

[0027] Figure 9 This is an internal structural diagram of a computer device in one embodiment;

[0028] Figure 10 This is a diagram of the internal structure of a computer device in another embodiment. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0030] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0031] As the power source for new energy vehicles, the operating status of the power battery during the charging process is directly related to the vehicle's safety, battery life and charging efficiency. If problems such as abnormal voltage drift, sudden temperature rise, or excessive current fluctuation occur during the charging process, and no timely warning or intervention measures are taken, it is very easy to cause safety accidents such as battery bulging, thermal runaway or even fire and explosion. At the same time, it will also accelerate battery aging and degradation and reduce the battery's cycle life.

[0032] In existing technologies, most methods rely on threshold-based rules to provide early warnings about the charging process of vehicle power batteries. However, existing charging warning methods suffer from low accuracy.

[0033] In view of this, this application provides a charging early warning method, which first acquires first charging operation data and predicted charging operation data of the vehicle's battery generated at multiple consecutive first charging moments, wherein the first charging period is earlier than the multiple first charging moments. Then, the predicted charging operation data is filtered based on the multiple first charging moments to obtain filtered predicted charging operation data. The actual operation data of the battery in the first charging period is added to the filtered predicted charging operation data to obtain new predicted charging operation data. Finally, based on the first charging operation data and the new predicted charging operation data, the second charging operation data of the battery at multiple consecutive second charging moments is determined, and the charging early warning information of the battery is determined based on the second charging operation data. The charging early warning method provided in this application predicts the charging operation data for multiple consecutive second charging moments by using first charging operation data generated at multiple consecutive first charging moments and predicted charging operation data for first charging periods earlier than multiple first charging moments. In other words, it uses the charging operation data generated in the current charging process and the charging operation data of historical charging processes to predict the charging operation data for the remaining stages of the current charging process. Since the predicted charging operation data is constantly updated, it takes into account both long-term charging patterns and adapts to short-term charging states, effectively improving the accuracy of the predicted charging operation data, and thus improving the accuracy of the charging early warning.

[0034] The charging early warning method provided in this application can be executed by a computer device, which can be a terminal, such as a vehicle's electronic control unit or an in-vehicle intelligent computing platform. The computer device can also be a server, which can communicate with the vehicle via a network and obtain charging operation data of the battery during the current charging process to predict the charging operation data of the vehicle's battery.

[0035] In one exemplary embodiment, such as Figure 1 As shown, a charging warning method is provided, which includes the following steps:

[0036] Step 101: Obtain the first charging operation data and the predicted charging operation data of the first charging period generated by the vehicle's battery at multiple consecutive first charging moments.

[0037] Optionally, if the process of the battery from the start of charging to the end of charging is defined as a charging process, then multiple consecutive first charging moments can be the charging moments that have already occurred in the current charging process.

[0038] The first charging period can be earlier than multiple first charging moments, and the first charging period can be a charging process that precedes the current charging process.

[0039] Predictive charging operation data can be the charging operation data for predicting the charging operation of the battery during the first charging period.

[0040] For example, the first charging operation data and the predicted charging operation data may include at least one of the following: battery state of charge, maximum value of individual cell voltage, minimum value of individual cell voltage, average value of individual cell voltage, maximum temperature of battery pack, minimum temperature of battery pack, charging current, individual cell voltage difference, and abnormal voltage deviation value.

[0041] In some exemplary embodiments, the computer device can acquire first charging operation data generated by the vehicle's battery at multiple consecutive first charging moments. Specifically, the computer device can acquire the first charging operation data generated by the vehicle's battery at multiple consecutive first charging moments through an on-board battery management system.

[0042] Furthermore, the computer equipment can also acquire predicted charging operation data of the vehicle's battery during the first charging period.

[0043] Specifically, computer equipment can obtain predicted charging operation data of the vehicle's battery during the first charging period from the vehicle's on-board storage unit, edge computing node, or cloud server.

[0044] Step 102: Filter the predicted charging operation data based on multiple first charging times to obtain filtered predicted charging operation data, and add the actual operation data of the battery in the first charging period to the filtered predicted charging operation data to obtain new predicted charging operation data.

[0045] In some exemplary embodiments, after acquiring first charging operation data and predicted charging operation data of the vehicle's battery at multiple consecutive first charging moments, the computer device can filter the predicted charging operation data based on the multiple first charging moments to obtain filtered predicted charging operation data.

[0046] Specifically, the computer equipment can filter out the earliest charging operation data corresponding to the earliest charging process from the predicted charging operation data based on multiple first charging moments.

[0047] In one possible implementation, the computer device can determine the time difference between the collection timestamp of each charging operation data in the predicted charging operation data and the same charging moment among multiple first charging moments, and identify the charging operation data with the largest time difference as the charging operation data corresponding to the earliest charging process for elimination.

[0048] Furthermore, after obtaining the filtered predicted charging operation data, the computer equipment can add the actual operation data of the battery during the first charging period to the filtered predicted charging operation data to obtain new predicted charging operation data.

[0049] Specifically, the computer equipment can first obtain the actual operating data of the battery during the first charging period, and then add the actual operating data to the filtered and processed predicted charging operating data to obtain new predicted charging operating data.

[0050] The following explanation will use the current charging process as the Nth charging process of the battery as an example to illustrate how to obtain new predictive charging operation data.

[0051] First, the computer device can obtain the charging operation data generated by the battery during the Nth charging process, which is the first charging operation data generated by the battery at multiple consecutive first charging moments. It can also obtain the historical charging operation data for predicting the charging operation of the battery during the (N-1)th charging process, which is the predicted charging operation data for the first charging period. This predicted charging operation data can include the charging operation data corresponding to the (N-5), (N-4), (N-3), and (N-2)th charging processes of the battery.

[0052] Next, the computer equipment can filter out the charging operation data corresponding to the earliest charging process from the charging operation data based on multiple first charging moments, that is, the charging operation data corresponding to the N-5th charging process of the battery, and then remove it.

[0053] Then, the computer device can obtain the actual operating data corresponding to the N-1th charging process of the battery and add it to the filtered predicted charging operating data to obtain new predicted charging operating data, which includes the charging operating data corresponding to the N-4th, N-3rd, N-2nd and N-1th charging processes of the battery.

[0054] Step 103: Based on the first charging operation data and the new predicted charging operation data, determine the second charging operation data of the battery at multiple consecutive second charging moments, and determine the charging warning information of the battery based on the second charging operation data.

[0055] For example, the second charging operation data is the predicted charging operation data of the battery during the remaining charging phase of the current charging process. The second charging operation data may include at least one of the following: battery state of charge sequence for the remaining charging phase, predicted maximum / minimum / average values ​​of individual cell voltages, predicted maximum / minimum temperatures of the battery pack, predicted charging current, predicted individual cell voltage differences, and predicted voltage anomaly deviations.

[0056] Charging warning information can include warnings of abnormal voltage drift, sudden temperature rise, excessive current fluctuation, and risk of thermal runaway during battery charging.

[0057] In some exemplary embodiments, after obtaining new predicted charging operation data, the computer device can determine the second charging operation data of the battery at a series of consecutive second charging moments based on the first charging operation data and the new predicted charging operation data.

[0058] Specifically, the computer equipment can first perform data preprocessing on the first charging operation data and the new predicted charging operation data to obtain preprocessed first charging operation data and new predicted charging operation data. Then, based on the preprocessed first charging operation data and the new predicted charging operation data, the second charging operation data of the battery at multiple consecutive second charging moments can be determined. Data preprocessing may include, but is not limited to, missing data completion, outlier removal, and standardization.

[0059] Furthermore, after obtaining the second charging operation data of the battery at multiple consecutive second charging moments, the computer equipment can determine the battery charging warning information based on the second charging operation data.

[0060] Specifically, the computer equipment can input the second charging operation data into a pre-trained charging warning recognition model to obtain the battery charging warning information output by the charging warning recognition model.

[0061] The aforementioned charging warning method first acquires the first charging operation data and predicted charging operation data of the vehicle's battery generated at multiple consecutive first charging moments, where the first charging period is earlier than the multiple first charging moments. Then, the predicted charging operation data is filtered based on the multiple first charging moments to obtain filtered predicted charging operation data. The actual operation data of the battery at the first charging period is added to the filtered predicted charging operation data to obtain new predicted charging operation data. Finally, based on the first charging operation data and the new predicted charging operation data, the second charging operation data of the battery at multiple consecutive second charging moments is determined, and the charging warning information of the battery is determined based on the second charging operation data. The charging early warning method provided in this application predicts the charging operation data for multiple consecutive second charging moments by using first charging operation data generated at multiple consecutive first charging moments and predicted charging operation data for first charging periods earlier than multiple first charging moments. In other words, it uses the charging operation data generated in the current charging process and the charging operation data of historical charging processes to predict the charging operation data for the remaining stages of the current charging process. Since the predicted charging operation data is constantly updated, it takes into account both long-term charging patterns and adapts to short-term charging states, effectively improving the accuracy of the predicted charging operation data, and thus improving the accuracy of the charging early warning.

[0062] In one exemplary embodiment, such as Figure 2 As shown, based on the first charging operation data and the new predicted charging operation data, the second charging operation data of the battery at multiple consecutive second charging moments is determined, including the following steps:

[0063] Step 201: Based on the preset state of charge increment, perform alignment and resampling processing on the first charging operation data and the new predicted charging operation data to obtain the target charging operation data.

[0064] Optionally, the preset state of charge (SOC) increment refers to a fixed step size for changing the state of charge that is pre-set according to actual needs. For example, 0.5% SOC / step, 1% SOC / step, or 2% SOC / step. In one optional embodiment, 1% SOC / step is used.

[0065] For example, the battery SOC growth rate varies under different charging scenarios, such as fast charging, slow charging, low-temperature charging, room-temperature charging, new battery charging, and aging battery charging. Furthermore, the original sampling time interval is irregular due to factors such as the acquisition frequency and communication delay, which affects the accuracy of the prediction. However, by performing aligned resampling based on a fixed SOC increment, different charging processes can be mapped to a unified physical dimension, eliminating modeling bias caused by time domain inconsistencies and improving the accuracy of the prediction.

[0066] In some exemplary embodiments, after obtaining the first charging operation data and the new predicted charging operation data, the computer device can perform alignment resampling processing on the first charging operation data and the new predicted charging operation data based on a preset state of charge increment to obtain the target charging operation data.

[0067] Specifically, the computer equipment can perform aligned resampling processing on the first charging operation data and the new predicted charging operation data with the state of charge increment as the step size, so as to obtain the aligned resampling first charging operation data and the new predicted charging operation data, and determine the aligned resampling first charging operation data and the new predicted charging operation data as the target charging operation data.

[0068] The above-mentioned alignment and resampling process is performed on the first charging operation data and the new predicted charging operation data based on the preset state of charge increment. By using a fixed increment, a unified representation of different charging processes is achieved, which solves the problem of charging trajectory misalignment in traditional time domain modeling and thus improves the accuracy of charging early warning.

[0069] Step 202: Input the target charging operation data into the pre-trained prediction model to obtain the second charging operation data output by the prediction model.

[0070] Optionally, the prediction model can be a long-term prediction model built based on an attention fusion mechanism. For example, an improved Transformer model, an attention-based Long Short-Term Memory (LSTM) network model, etc.

[0071] For example, the prediction model can be trained end-to-end. The training data for the prediction model can be real electric vehicle charging operation data, and the loss function during the training process can be the mean squared error between the predicted and actual values. During the training process, the model parameters of the prediction model can be iteratively updated by an optimizer to minimize the prediction error of various battery state variables within the future SOC range. Simultaneously, the prediction model also supports online incremental training. After each charging cycle, the actual operation data of that charging cycle can be added to the training set to fine-tune the model parameters and adapt them to the characteristics of the entire battery lifecycle.

[0072] In some exemplary embodiments, after obtaining the target charging operation data, the computer device can input the target charging operation data into a pre-trained prediction model to obtain the second charging operation data output by the prediction model.

[0073] In one exemplary embodiment, such as Figure 3As shown, the target charging operation data is obtained by aligning and resampling the first charging operation data and the new predicted charging operation data based on the preset state of charge increment, including the following steps:

[0074] Step 301: Using the state of charge increment as the step size, sample the first charging operation data and the new predicted charging operation data at equal intervals to obtain the aligned first charging operation data and the new predicted charging operation data.

[0075] In some exemplary embodiments, the computer device may sample the first charging operation data and the new predicted charging operation data at equal intervals with the state of charge increment as the step size, so as to obtain the aligned first charging operation data and the new predicted charging operation data.

[0076] Specifically, the computer equipment can first determine the state of charge (SOC) data and other charging operation parameters corresponding to the SOC data in the first charging operation data and the new predicted charging operation data. Then, based on the SOC data and the other charging operation parameters corresponding to the SOC data, it can construct an associated dataset of SOC data and other charging operation parameters. Then, with a preset SOC increment as a fixed step size, it can generate a continuous and equally spaced sequence of SOC sampling points. For each SOC sampling point in the sequence of SOC sampling points, it can match the corresponding other charging operation parameters from the associated dataset of SOC data and other charging operation parameters to obtain the first charging operation data and the new predicted charging operation data corresponding to each SOC sampling point.

[0077] For example, with a preset state of charge increment of 1% SOC / step, the computer device can generate a sequence of state of charge sampling points of 0%, 1%, 2%...100%. For the 1% state of charge sampling point, the computer can match the maximum, minimum, and average values ​​of the individual battery voltage, the highest and lowest temperatures of the battery pack, the charging current, and the individual battery voltage difference, etc., from the state of charge data of the first charging operation data and the new predicted charging operation data.

[0078] Furthermore, if a certain state of charge sampling point cannot be matched with the corresponding charging operation parameters in the first charging operation data and the new predicted charging operation data, interpolation can be performed based on the charging operation parameters corresponding to the adjacent state of charge sampling points to obtain the first charging operation data and the new predicted charging operation data corresponding to each state of charge sampling point.

[0079] Step 302: Determine the aligned first charging operation data and the new predicted charging operation data as the target charging operation data.

[0080] In some exemplary embodiments, after obtaining aligned first charging operation data and new predicted charging operation data, the computer device can determine the aligned first charging operation data and the new predicted charging operation data as target charging operation data.

[0081] In one exemplary embodiment, such as Figure 4 As shown, the prediction model includes an encoder, which comprises a first low-rank attention branch module, a first linear attention branch module, and a first fusion module. The target charging operation data is input into the pre-trained prediction model for prediction processing to obtain the second charging operation data, including the following steps:

[0082] Step 401: Input the target charging operation data into the first low-rank attention branch module and the first linear attention branch module of the encoder to obtain the long-term evolution characteristics output by the first low-rank attention branch module and the local continuous characteristics output by the first linear attention branch module.

[0083] Optionally, long-term evolution features can be used to reflect long-term effects such as battery voltage drift and inconsistency accumulation; local continuous features can be used to reflect the smooth change trend of parameters such as battery voltage and battery temperature within adjacent SOC ranges.

[0084] In some exemplary embodiments, the computer device can input the target charging operation data into the first low-rank attention branch module and the first linear attention branch module of the encoder, respectively, to obtain the long-term evolution features output by the first low-rank attention branch module and the local continuous features output by the first linear attention branch module.

[0085] Specifically, the computer device can input new predicted charging operation data from the target charging operation data into the first low-rank attention branch module of the encoder. The first low-rank attention branch module calculates attention weights in the low-dimensional subspace by performing low-dimensional projection on the query vector, key vector, and value vector, and models the global dependency of the new predicted charging operation data to obtain long-term evolutionary features. The computer device can also input new predicted charging operation data from the target charging operation data into the first linear attention branch module of the encoder. The first linear attention branch module reconstructs the attention calculation process in the form of a kernel function, making the attention calculation complexity approximately linear, and models the local continuous change trend of the new predicted charging operation data to obtain local continuous features.

[0086] Step 402: Input the long-term evolution features and local continuous features into the first fusion module of the encoder to obtain the global charging feature information output by the first fusion module.

[0087] Optionally, the global charging characteristic information can be a set of features that reflect the long-term characteristics of battery charging.

[0088] In some exemplary embodiments, after obtaining long-term evolution features and local continuous features, the computer device can input the long-term evolution features and local continuous features into the first fusion module of the encoder to obtain the global charging feature information output by the first fusion module.

[0089] Specifically, the computer device can first perform feature concatenation processing on long-term evolutionary features and local continuous features to obtain the first fused input features, and then input the first fused input features into the first fusion module of the encoder. The first fusion module has at least one layer of fully connected network and a sigmoid activation function.

[0090] Furthermore, the first fusion module first performs dimensional transformation and feature mapping on the first fusion input features through a fully connected network, and then uses the Sigmoid activation function to dynamically calculate the first fusion weights corresponding to the long-term evolution features and local continuous features based on the feature distribution of the first fusion input features. The first fusion weights can take values ​​ranging from 0 to 1.

[0091] Furthermore, the first fusion module can perform weighted fusion processing on long-term evolution features and local continuous features based on the first fusion weight to obtain global charging feature information.

[0092] The adaptive fusion of low-rank attention and linear attention achieves a dynamic balance between global long-term evolution characteristics and local continuous change characteristics. It takes into account long-term effects such as battery voltage drift and inconsistency accumulation, as well as the smooth change trend of parameters in adjacent SOC intervals. This avoids the omission of global or local information by a single attention mechanism, thereby effectively improving the accuracy of charging warning.

[0093] Step 403: Determine the second charging operation data based on global charging feature information and target charging operation data.

[0094] In some exemplary embodiments, after obtaining global charging feature information, the computer device can determine second charging operation data based on the global charging feature information and the target charging operation data.

[0095] In one exemplary embodiment, such as Figure 5 As shown, the prediction model also includes a decoder, which comprises a second low-rank attention branch module, a second linear attention branch module, and a second fusion module. Based on global charging feature information and target charging operation data, the decoder determines the second charging operation data, including the following steps:

[0096] Step 501: Input the target charging operation data and global charging feature information into the second low-rank attention branch module and the second linear attention branch module of the decoder to obtain the long-term trend features output by the second low-rank attention branch module and the local evolution features output by the second linear attention branch module.

[0097] Optionally, long-term trend features can be used to reflect the overall development trend of the battery state as a function of SOC during the remaining charging phase. Local evolution features can be used to reflect the continuous fluctuations and detailed evolution characteristics of the battery state during the remaining charging phase.

[0098] In some exemplary embodiments, the computer device can input target charging operation data and global charging feature information into the second low-rank attention branch module and the second linear attention branch module of the decoder to obtain the long-term trend features output by the second low-rank attention branch module and the local evolution features output by the second linear attention branch module.

[0099] Specifically, the computer device can input the first charging operation data and global charging feature information from the target charging operation data into the second low-rank attention branch module of the decoder. The second low-rank attention branch module can perform low-dimensional projection on the query vector, key vector, and value vector, calculate the attention weight in the low-dimensional subspace, and combine the long-term charging pattern of the battery in the global charging feature information with the real-time state of the first charging operation data to model the global dependency relationship of the remaining stage of the current charging process to obtain long-term trend features. The computer device can also input the first charging operation data and global charging feature information into the second linear attention branch module of the decoder. The second linear attention branch module can reconstruct the attention calculation process in the form of kernel function to make the attention calculation complexity approximately linear. It combines the common pattern of local battery charging in the global charging feature information with the parameter changes of adjacent SOC nodes in the first charging operation data to obtain local evolution features.

[0100] Step 502: Input the long-term trend features and local evolution features into the second fusion module of the decoder to obtain the second charging operation data output by the second fusion module.

[0101] In some exemplary embodiments, after obtaining long-term trend features and local evolution features, the computer device can input the long-term trend features and local evolution features into the second fusion module of the decoder to obtain the second charging operation data output by the second fusion module.

[0102] Specifically, the computer device can first perform feature concatenation processing on long-term trend features and local evolution features to obtain a second fused input feature, and then input the second fused input feature into the second fusion module of the decoder. The second fusion module incorporates at least one layer of fully connected network and a sigmoid activation function.

[0103] Furthermore, the second fusion module can first perform dimensional transformation and feature mapping on the second fusion input features through a fully connected network, and then use the Sigmoid activation function to dynamically calculate the second fusion weights corresponding to the long-term trend features and local evolution features based on the feature distribution of the second fusion input features. The value range of the second fusion weights can be 0-1.

[0104] Furthermore, the second fusion module can perform weighted fusion processing on long-term trend features and local evolution features based on the second fusion weight, so as to output the predicted charging operation data within the remaining SOC range of the current charging process.

[0105] In one exemplary embodiment, such as Figure 6 As shown, determining battery charging warning information based on the second charging operation data includes the following steps:

[0106] Step 601: Obtain a preset set of charging fault thresholds.

[0107] Optionally, the charging fault threshold set includes charging fault thresholds for at least two different risk levels corresponding to multiple charging fault types.

[0108] For example, multiple charging fault types may include voltage difference fault, voltage abnormality fault, and temperature abnormality fault, and at least two different risk levels can be set for each of the multiple charging fault types.

[0109] For example, voltage difference faults can be set with a first-level warning threshold and a second-level warning threshold; voltage outlier faults can be set with a graded threshold based on the difference between the lowest voltage and the average voltage of a single cell; and temperature abnormal faults can be set with a warning threshold and an emergency threshold based on the highest temperature of the battery pack.

[0110] Step 602: If the second charging operation data is greater than the corresponding charging fault threshold in the charging fault threshold set, determine the battery charging warning information according to the risk level and charging fault type corresponding to the charging fault threshold.

[0111] In an exemplary embodiment, after obtaining a preset set of charging fault thresholds, the computer device can determine the battery charging warning information based on the risk level and charging fault type corresponding to the charging fault threshold if the second charging operation data is greater than the corresponding charging fault threshold in the set of charging fault thresholds.

[0112] Specifically, the computer equipment can first divide the second charging operation data into multiple future SOC intervals, and extract the corresponding predicted charging operation parameters in each SOC interval, which may include the difference between the maximum and minimum values ​​of the single cell voltage, the difference between the minimum and average voltage of the single cell, the maximum temperature of the battery pack, the minimum temperature of the battery pack, etc.

[0113] Furthermore, the predicted charging operation parameters for each SOC range are compared with the corresponding grading thresholds for charging fault types in the charging fault threshold set. That is, the difference between the maximum and minimum voltage of a single cell is compared with the multi-level threshold for voltage difference anomaly faults, the difference between the minimum and average voltage of a single cell is compared with the grading threshold for voltage outlier anomaly faults, and the highest and lowest temperatures of the battery pack are compared with the risk level threshold for temperature anomaly faults.

[0114] Furthermore, it is determined whether the predicted charging operating parameters of a certain SOC range reach or exceed the charging fault threshold corresponding to a certain level. If the predicted charging operating parameters of a certain SOC range are detected to meet the condition, the charging fault type and risk level corresponding to the charging fault threshold are matched, and the charging warning information of the battery is determined by combining the SOC range that triggers the warning. The charging warning information may include the fault type, risk level and the SOC range that triggers the warning.

[0115] In one exemplary embodiment, such as Figure 7 As shown, another charging warning method is provided, which includes the following steps:

[0116] Step 701: Obtain the first charging operation data and the predicted charging operation data of the first charging period generated by the vehicle's battery at multiple consecutive first charging moments; the first charging period is earlier than multiple first charging moments.

[0117] Step 702: Based on multiple first charging times, filter the predicted charging operation data to obtain the filtered predicted charging operation data, and add the actual operation data of the battery in the first charging period to the filtered predicted charging operation data to obtain new predicted charging operation data.

[0118] Step 703: Using the state of charge increment as the step size, sample the first charging operation data and the new predicted charging operation data at equal intervals to obtain the aligned first charging operation data and the new predicted charging operation data; determine the aligned first charging operation data and the new predicted charging operation data as the target charging operation data;

[0119] Step 704: Input the target charging operation data into the first low-rank attention branch module and the first linear attention branch module of the encoder to obtain the long-term evolution features output by the first low-rank attention branch module and the local continuous features output by the first linear attention branch module; input the long-term evolution features and the local continuous features into the first fusion module of the encoder to obtain the global charging feature information output by the first fusion module.

[0120] Step 705: Input the target charging operation data and global charging feature information into the second low-rank attention branch module and the second linear attention branch module of the decoder to obtain the long-term trend features output by the second low-rank attention branch module and the local evolution features output by the second linear attention branch module; input the long-term trend features and local evolution features into the second fusion module of the decoder to obtain the second charging operation data output by the second fusion module;

[0121] Step 706: Obtain a preset set of charging fault thresholds, which includes at least two different risk levels of charging fault thresholds corresponding to multiple charging fault types; if the second charging operation data is greater than the corresponding charging fault threshold in the set of charging fault thresholds, determine the battery charging warning information according to the risk level and charging fault type corresponding to the charging fault threshold.

[0122] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0123] Based on the same inventive concept, this application also provides a charging warning device for implementing the charging warning method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more embodiments of the charging warning device provided below can be found in the limitations of the charging warning method described above, and will not be repeated here.

[0124] In one exemplary embodiment, such as Figure 8 As shown, a charging warning device 800 is provided, including: an acquisition module 801, an execution module 802, and a determination module 803, wherein:

[0125] The acquisition module 801 is used to acquire first charging operation data and predicted charging operation data of the vehicle's battery generated at multiple consecutive first charging moments; the first charging period is earlier than the multiple first charging moments.

[0126] The execution module 802 is used to filter the predicted charging operation data based on multiple first charging times to obtain the filtered predicted charging operation data, and add the actual operation data of the battery in the first charging period to the filtered predicted charging operation data to obtain new predicted charging operation data.

[0127] The determination module 803 is used to determine the second charging operation data of the battery at multiple consecutive second charging moments based on the first charging operation data and the new predicted charging operation data, and to determine the charging warning information of the battery based on the second charging operation data.

[0128] In one embodiment, the determining module 803 is specifically used to perform aligned resampling processing on the first charging operation data and the new predicted charging operation data based on the preset state of charge increment to obtain the target charging operation data; and input the target charging operation data into the pre-trained prediction model for prediction processing to obtain the second charging operation data.

[0129] In one embodiment, the determining module 803 is specifically used to sample the first charging operation data and the new predicted charging operation data at equal intervals with the state of charge increment as the step size, to obtain the aligned first charging operation data and the new predicted charging operation data; and to determine the aligned first charging operation data and the new predicted charging operation data as the target charging operation data.

[0130] In one embodiment, the prediction model includes an encoder, which includes a first low-rank attention branch module, a first linear attention branch module, and a first fusion module. A determination module 803 is specifically used to input target charging operation data into the first low-rank attention branch module and the first linear attention branch module of the encoder to obtain long-term evolution features output by the first low-rank attention branch module and local continuous features output by the first linear attention branch module; input the long-term evolution features and local continuous features into the first fusion module of the encoder to obtain global charging feature information output by the first fusion module; and determine second charging operation data based on the global charging feature information and the target charging operation data.

[0131] In one embodiment, the prediction model further includes a decoder, which includes a second low-rank attention branch module, a second linear attention branch module, and a second fusion module; the determination module 803 is specifically used to input the target charging operation data and global charging feature information into the second low-rank attention branch module and the second linear attention branch module of the decoder to obtain the long-term trend features output by the second low-rank attention branch module and the local evolution features output by the second linear attention branch module; and input the long-term trend features and the local evolution features into the second fusion module of the decoder to obtain the second charging operation data output by the second fusion module.

[0132] In one embodiment, the determining module 803 is specifically used to obtain a preset set of charging fault thresholds, which includes at least two charging fault thresholds with different risk levels corresponding to multiple charging fault types; when the second charging operation data is greater than the corresponding charging fault threshold in the set of charging fault thresholds, the charging warning information of the battery is determined according to the risk level and charging fault type corresponding to the charging fault threshold.

[0133] Each module in the aforementioned charging warning device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0134] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a charging warning method.

[0135] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 10As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When executed by the processor, the computer program implements a charging warning method.

[0136] Those skilled in the art will understand that Figure 9 and Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0137] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the charging warning method described in any of the above embodiments.

[0138] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the charging warning method described in any of the above embodiments.

[0139] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the charging warning method described in any of the above embodiments.

[0140] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0141] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0142] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A charging early warning method, characterized in that, The method includes: Acquire first charging operation data and predicted charging operation data for a first charging period generated by the vehicle's battery at multiple consecutive first charging moments; the first charging period is earlier than the multiple first charging moments. Based on the multiple first charging times, the predicted charging operation data is filtered to obtain filtered predicted charging operation data. The actual operation data of the battery during the first charging period is added to the filtered predicted charging operation data to obtain new predicted charging operation data. Based on the first charging operation data and the new predicted charging operation data, the second charging operation data of the battery at multiple consecutive second charging moments is determined, and the charging warning information of the battery is determined based on the second charging operation data.

2. The method according to claim 1, characterized in that, The step of determining the second charging operation data of the battery at multiple consecutive second charging moments based on the first charging operation data and the new predicted charging operation data includes: Based on the preset state of charge increment, the first charging operation data and the new predicted charging operation data are aligned and resampled to obtain the target charging operation data. The target charging operation data is input into a pre-trained prediction model for prediction processing to obtain the second charging operation data.

3. The method according to claim 2, characterized in that, The first charging operation data and the new predicted charging operation data are aligned and resampled based on a preset state of charge increment to obtain target charging operation data, including: Using the state of charge increment as the step size, the first charging operation data and the new predicted charging operation data are sampled at equal intervals to obtain the aligned first charging operation data and the new predicted charging operation data. The aligned first charging operation data and the new predicted charging operation data are determined as the target charging operation data.

4. The method according to claim 2, characterized in that, The prediction model includes an encoder, which includes a first low-rank attention branch module, a first linear attention branch module, and a first fusion module; the step of inputting the target charging operation data into the pre-trained prediction model for prediction processing to obtain the second charging operation data includes: The target charging operation data is input into the first low-rank attention branch module and the first linear attention branch module of the encoder to obtain the long-term evolution features output by the first low-rank attention branch module and the local continuous features output by the first linear attention branch module. The long-term evolution features and the local continuous features are input into the first fusion module of the encoder to obtain the global charging feature information output by the first fusion module; The second charging operation data is determined based on the global charging feature information and the target charging operation data.

5. The method according to claim 4, characterized in that, The prediction model also includes a decoder, which includes a second low-rank attention branch module, a second linear attention branch module, and a second fusion module. The step of determining the second charging operation data based on the global charging feature information and the first charging operation data in the target charging operation data includes: The target charging operation data and the global charging feature information are input into the second low-rank attention branch module and the second linear attention branch module of the decoder to obtain the long-term trend features output by the second low-rank attention branch module and the local evolution features output by the second linear attention branch module. The long-term trend features and the local evolution features are input into the second fusion module of the decoder to obtain the second charging operation data output by the second fusion module.

6. The method according to any one of claims 1 to 5, characterized in that, The step of determining the battery charging warning information based on the second charging operation data includes: Obtain a preset set of charging fault thresholds, wherein the set of charging fault thresholds includes at least two charging fault thresholds of different risk levels corresponding to multiple charging fault types; If the second charging operation data is greater than the corresponding charging fault threshold in the charging fault threshold set, the charging warning information of the battery is determined according to the risk level and charging fault type corresponding to the charging fault threshold.

7. A charging early warning device, characterized in that, The device includes: The acquisition module is used to acquire first charging operation data and predicted charging operation data of the vehicle's battery during multiple consecutive first charging moments; the first charging period is earlier than the multiple first charging moments. The execution module is used to filter the predicted charging operation data based on the plurality of first charging times to obtain filtered predicted charging operation data, and to add the actual operation data of the battery in the first charging period to the filtered predicted charging operation data to obtain new predicted charging operation data. The determination module is used to determine the second charging operation data of the battery at multiple consecutive second charging moments based on the first charging operation data and the new predicted charging operation data, and to determine the charging warning information of the battery based on the second charging operation data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images