A method and system for predicting lithium battery failures

Abstract

The present invention relates to a lithium battery fault prediction method and system, belonging to the field of lithium battery fault diagnosis, including: identifying the similarity between the actual condition graph and the standard condition graph, and calculating to obtain the condition saliency; according to a plurality of voltage change saliencies, a plurality of current change saliencies, and a plurality of temperature change saliencies, performing overcharge fault prediction on the lithium battery pack to obtain a first overcharge fault coefficient, and using the condition saliency to perform correction calculation on the first overcharge fault coefficient to obtain a second overcharge fault coefficient, which is used as the lithium battery fault prediction result. Through the present application, the technical problem that the existing lithium battery overcharge fault prediction methods usually rely on fixed thresholds or simple analyses based on single parameters, resulting in insufficient prediction accuracy and reliability, can be solved; by fusing multi-dimensional data for overcharge fault prediction, the accuracy and reliability of fault prediction can be improved, so that potential risks during the charging process can be identified in advance, and the use safety of lithium batteries can be significantly improved.

Description

Technical Field

[0001] The present invention relates to the field of lithium battery fault diagnosis, and particularly to a method and system for predicting lithium battery faults. Background Art

[0002] With the wide application of lithium batteries in fields such as electric vehicles, energy storage systems, and consumer electronics, the safety and reliability of batteries have become important concerns during the design and use processes. During the charging process of lithium batteries, if overcharging occurs, it may lead to battery damage, performance degradation, and even safety accidents such as fires. Therefore, accurately and real-time predicting overcharging faults of lithium batteries is crucial for improving the safety of battery packs and extending their service life.

[0003] Currently, existing methods for predicting overcharging faults of lithium batteries usually rely on fixed thresholds or simple analyses based on single parameters. For example, if the battery voltage exceeds 4.2V, the system immediately considers overcharging to have occurred and issues an alarm; or when the temperature of the battery exceeds a certain set value, it is considered that the battery may have a risk of overcharging, etc. Lack of flexibility and adaptability results in insufficient accuracy and reliability of fault prediction. Summary of the Invention

[0004] Aiming at the technical problems that existing methods for predicting overcharging faults of lithium batteries usually rely on fixed thresholds or simple analyses based on single parameters, resulting in insufficient accuracy and reliability of prediction, the present invention provides a method and system for predicting lithium battery faults to solve this problem.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] In a first aspect, the present invention provides a method for predicting lithium battery faults, including: during the charging process of a lithium battery pack, collecting the voltages, currents, and temperatures of multiple lithium battery cells in the lithium battery pack to form multiple voltage sequences, multiple current sequences, and multiple temperature sequences, and obtaining the standard voltage sequence, standard current sequence, and standard temperature sequence of the lithium battery cell charging, and performing significant change analysis to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances; obtaining the standard usage condition parameters of the lithium battery pack, constructing a standard condition graph, and collecting the actual usage condition parameters of the current lithium battery pack, constructing an actual condition graph, wherein, according to the multiple temperature change significances, performing consistency analysis on the lithium battery pack to obtain consistency condition parameters, and constructing the actual condition graph; identifying the similarity between the actual condition graph and the standard condition graph, and calculating to obtain a condition significance; according to the multiple voltage change significances, multiple current change significances, and multiple temperature change significances, performing overcharging fault prediction on the lithium battery pack to obtain a first overcharging fault coefficient, and using the condition significance to perform correction calculation on the first overcharging fault coefficient to obtain a second overcharging fault coefficient as the lithium battery fault prediction result.

[0007] In a second aspect, the present invention provides a lithium battery fault prediction system, including: a change significance analysis module, configured to collect voltages, currents, and temperatures of a plurality of lithium battery cells in a lithium battery pack during the charging process of the lithium battery pack, form a plurality of voltage sequences, a plurality of current sequences, and a plurality of temperature sequences, and obtain a standard voltage sequence, a standard current sequence, and a standard temperature sequence for the charging of the lithium battery cells, and perform change significance analysis to obtain a plurality of voltage change significances, a plurality of current change significances, and a plurality of temperature change significances; an actual condition map construction module, configured to obtain standard usage condition parameters of the lithium battery pack, construct a standard condition map, and collect actual usage condition parameters of the current lithium battery pack, and construct an actual condition map, wherein, according to the plurality of temperature change significances, consistency analysis of the lithium battery pack is performed to obtain consistency condition parameters, and the actual condition map is constructed; a condition significance calculation module, configured to identify the similarity between the actual condition map and the standard condition map, and calculate to obtain a condition significance; an overcharge fault prediction module, configured to perform overcharge fault prediction on the lithium battery pack according to the plurality of voltage change significances, the plurality of current change significances, and the plurality of temperature change significances, obtain a first overcharge fault coefficient, and perform correction calculation on the first overcharge fault coefficient by using the condition significance to obtain a second overcharge fault coefficient, which is used as the lithium battery fault prediction result.

[0008] In a third aspect, the present invention further provides an electronic device, including:

[0009] at least one processor; a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the steps of the method according to any one of the above first aspects.

[0010] In a fourth aspect, a computer-readable storage medium stores a computer program, and when the computer program is executed, it implements the steps of the method according to any one of the above first aspects.

[0011] The beneficial effects of the present invention are as follows: During the charging process of the lithium battery pack, the voltages, currents, and temperatures of multiple lithium battery cells within the lithium battery pack are collected to form multiple voltage sequences, multiple current sequences, and multiple temperature sequences. Moreover, the standard voltage sequence, standard current sequence, and standard temperature sequence for the charging of the lithium battery cells are obtained, and significant change analysis is carried out to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances. On the other hand, the standard usage condition parameters of the lithium battery pack are obtained to construct a standard condition diagram, and the actual usage condition parameters of the current lithium battery pack are collected to construct an actual condition diagram. Among them, according to the multiple temperature change significances, consistency analysis of the lithium battery pack is performed to obtain consistency condition parameters for constructing the actual condition diagram. Then, the similarity between the actual condition diagram and the standard condition diagram is identified, and the condition significance is calculated. Further, based on the multiple voltage change significances, multiple current change significances, and multiple temperature change significances, overcharge fault prediction of the lithium battery pack is carried out to obtain a first overcharge fault coefficient. Finally, the condition significance is used to perform a correction calculation on the first overcharge fault coefficient to obtain a second overcharge fault coefficient as the lithium battery fault prediction result. That is to say, by fusing multi-dimensional data for overcharge fault prediction of lithium batteries, the accuracy and reliability of overcharge fault prediction can be improved, thereby potential risks during the charging process can be identified in advance, serious consequences caused by overcharge faults can be avoided, and the usage safety of lithium batteries can be significantly improved. Description of the Drawings

[0012] Figure 1 It is a schematic flowchart of a method for predicting lithium battery faults provided by the present invention;

[0013] Figure 2 It is a schematic structural diagram of a system for predicting lithium battery faults provided by the present invention;

[0014] Figure 3 It is a schematic structural diagram of an electronic device provided by the present invention;

[0015] Figure 4 It is a schematic structural diagram of a computer-readable storage medium provided by the present invention.

[0016] In the drawings, the components represented by each reference numeral are described as follows:

[0017] Significant change analysis module 01, actual condition diagram construction module 02, condition significance calculation module 03, overcharge fault prediction module 04, electronic device 500, memory 510, processor 520, first computer program 511, computer-readable storage medium 600, second computer program 611. Detailed Embodiments

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0019] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality of" means two or more, unless otherwise specifically defined.

[0020] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily construed as being more preferred or having more advantages than other embodiments. In order for any person skilled in the art to implement and use the present invention, the following description is given. In the following description, details are set forth for purposes of explanation. It should be understood that those of ordinary skill in the art can recognize that the present invention can be implemented without the use of these specific details. In other instances, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope that conforms to the principles and features disclosed in the present invention.

[0021] Embodiment 1, as Figure 1 shown, the embodiment of the present invention provides a method for predicting lithium battery faults, which specifically includes the following steps:

[0022] S100: During the charging process of the lithium battery pack, collect the voltages, currents, and temperatures of multiple lithium battery cells in the lithium battery pack to form multiple voltage sequences, multiple current sequences, and multiple temperature sequences, and obtain the standard voltage sequence, standard current sequence, and standard temperature sequence of the lithium battery cell charging, and perform a significant change analysis to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances.

[0023] Further, step S100 of this application further includes:

[0024] S110: During the charging process of the lithium battery pack, collect the voltages, currents, and temperatures of multiple lithium battery cells at multiple time nodes before the charge coefficient reaches the full charge threshold, and form multiple voltage sequences, multiple current sequences, and multiple temperature sequences; S120: Obtain the standard voltages, standard currents, and standard temperatures within the multiple time nodes, and construct a standard voltage sequence, a standard current sequence, and a standard temperature sequence; S130: Calculate the deviations between the multiple voltage sequences, multiple current sequences, and multiple temperature sequences and the standard voltage sequence, standard current sequence, and standard temperature sequence respectively, to obtain multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences.

[0025] Specifically, configure the data acquisition time interval, which can be set according to the fault prediction accuracy requirement. The higher the fault prediction accuracy requirement, the shorter the data acquisition time interval. For example, set the data acquisition time interval to 1 minute, that is, data acquisition is performed every 1 minute. Then, during the charging process of the lithium battery pack, collect the voltages, currents, and temperatures of multiple lithium battery cells at multiple time nodes before the charge coefficient reaches the full charge threshold. The state of charge (SOC) is a value that describes the ratio of the current battery charge to the maximum charge, usually expressed as a percentage. During the charging process, the SOC of the battery gradually increases from 0% and when the SOC reaches 100%, the battery is considered fully charged. During the charging process, the voltage, current, and temperature of the battery will change dynamically. Then, arrange the voltages, currents, and temperatures of multiple lithium battery cells at multiple time nodes in the order of the time nodes, and obtain multiple voltage sequences, multiple current sequences, and multiple temperature sequences. Among them, each lithium battery cell includes a voltage sequence, a current sequence, and a temperature sequence.

[0026] Next, obtain the standard voltages, standard currents, and standard temperatures within the multiple time nodes, and arrange the standard voltages, standard currents, and standard temperatures within the multiple time nodes in the order of the time nodes to obtain a standard voltage sequence, a standard current sequence, and a standard temperature sequence. Further, based on the standard voltage sequence, standard current sequence, and standard temperature sequence, perform deviation calculations on the multiple voltage sequences, multiple current sequences, and multiple temperature sequences respectively, that is, calculate the deviations between the actual voltage, current, temperature and the standard values in each lithium battery cell at the same time node, and obtain multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences. Among them, the voltage deviation reflects the difference between the actual voltage and the standard voltage of each lithium battery cell; the current deviation reflects the difference between the actual charging current and the standard current of each battery cell; the temperature deviation reflects the difference between the actual temperature and the standard temperature of each battery cell. By calculating the deviation sequences of voltage, current, and temperature, the state changes of each unit during the lithium battery charging process can be monitored in real time, and potential faults and abnormal situations can be identified in a timely manner.

[0027] S140: Based on the multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences, perform a significant change analysis to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances.

[0028] Furthermore, step S140 of the present application further includes:

[0029] S141: Select a first selected voltage deviation within the first voltage deviation sequence, and randomly select a first random voltage deviation, calculate the ratio of the first selected voltage deviation and the first random voltage deviation to obtain a first selected voltage change significance; S142: Continue to calculate and obtain multiple first selected voltage change significances, and calculate the mean value to obtain a first voltage change significance; S143: Continue to calculate and obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances based on the multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences.

[0030] Specifically, first, randomly select any one of the multiple voltage deviation sequences and set it as the first voltage deviation sequence, and select the voltage deviation at the first time node within the first voltage deviation sequence as the first selected voltage deviation, where the first time node refers to the time node with the highest ranking, that is, the first node; then randomly select the voltage deviation at any time node other than the first time node within the first voltage deviation sequence as the first random voltage deviation; further calculate the ratio of the first selected voltage deviation and the first random voltage deviation, and set the voltage deviation ratio as the first selected voltage change significance, where a larger significance value indicates that the selected voltage deviation has a greater change compared to the randomly selected voltage deviation, meaning there is an abnormal voltage fluctuation during the charging process or a risk of a large deviation in the battery.

[0031] Further, using the same method, continue to calculate the ratio of the first selected voltage deviation and other random voltage deviations to obtain multiple first selected voltage change significances, and calculate the mean value of the multiple first selected voltage change significances to obtain a first voltage change significance. Then continue to perform a significance calculation based on the multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances. Among them, the voltage change significance reflects the degree of battery voltage change of each lithium battery unit during the charging process, the current change significance reflects the degree of current change of each lithium battery unit during the charging process, and the temperature change significance reflects the degree of change of the battery temperature of each lithium battery unit during the charging process.

[0032] S200: Obtain the standard usage condition parameters of the lithium battery pack, construct a standard condition graph, and collect the actual usage condition parameters of the current lithium battery pack to construct an actual condition graph. Among them, perform consistency analysis on the lithium battery pack according to the multiple temperature change significances to obtain consistency condition parameters and construct the actual condition graph.

[0033] Further, step S200 of this application further includes:

[0034] S210: Obtain the standard usage condition parameters of the lithium battery pack, where the standard usage condition parameters include standard ambient temperature, standard charging voltage, and standard consistency parameter; S220: Set the standard ambient temperature, standard charging voltage, and standard consistency parameter as preset values and use them as the pixel values of the image data to construct a standard condition graph; S230: Collect the actual usage condition parameters of the current lithium battery pack, where the actual usage condition parameters include actual ambient temperature, actual charging voltage, and actual consistency parameter, and the actual consistency parameter is obtained by calculating the variance of the multiple temperature change significances; S240: Calculate the ratios of the actual ambient temperature, actual charging voltage, and actual consistency parameter to the standard ambient temperature, standard charging voltage, and standard consistency parameter, multiply by the preset value, and use them as the pixel values of the image data to construct an actual condition graph.

[0035] Specifically, first, obtain the standard usage condition parameters of the lithium battery pack, where the standard usage condition parameters include standard ambient temperature, standard charging voltage, and standard consistency parameter. The standard consistency parameter characterizes the uniformity of the temperature performance of each battery cell in the battery pack during the charging process, which can be represented by the temperature variance and is used to reflect the consistency of the temperature distribution among the battery cells.

[0036] Next, map the standard ambient temperature, standard charging voltage, and standard consistency parameter to the pixel values of the image data. For example, set the standard ambient temperature (such as 25°C) as the preset value of the gray value of a certain image area in the image, for example, the gray value 100, to represent this environmental condition; set the standard charging voltage (such as 4.2V) as the preset value of the pixel value in the image, and the same pixel value preset value (such as gray value 100) can be given to this standard charging voltage to represent the ideal voltage condition during the charging process; set the standard consistency parameter (temperature variance) to 0, indicating that the temperatures of all the cells in the battery pack are exactly the same, which is the best state. When the temperature variance is 0, it is mapped to another pixel value preset value (such as gray value 100), indicating that the consistency of the battery pack is very good. Based on the above-set pixel values, construct a standard condition graph, where the standard condition graph includes three regions, and each region represents a standard condition parameter respectively.

[0037] Then, collect the actual usage condition parameters of the current lithium battery pack. Among them, the actual usage condition parameters include the actual ambient temperature, the actual charging voltage, and the actual consistency parameter. The actual consistency parameter is obtained by calculating the variance of the multiple temperature change significances, that is, setting the variance of the multiple temperature change significances as the actual consistency parameter. Further calculate the ratio of the actual ambient temperature to the standard ambient temperature, denoted as the temperature parameter ratio, calculate the ratio of the actual charging voltage to the standard charging voltage and denote it as the voltage parameter ratio, and calculate the ratio of the actual consistency parameter to the standard consistency parameter and denote it as the consistency parameter ratio; and multiply the temperature parameter ratio, the voltage parameter ratio, and the consistency parameter ratio by the pixel value preset value (such as the gray value 100) as the pixel value of the image data to construct the actual condition map. Among them, the gray value (i.e., the pixel value) of each region in the image represents the difference between the actual conditions and the standard conditions of the battery pack during the charging process. When the actual ambient temperature, voltage, and consistency parameter of the battery are quite different from the standard conditions, the difference in the pixel values (i.e., the gray values) between the actual condition map and the standard condition map is large, and obvious changes will occur in the image, indicating that there may be abnormalities or faults during the battery charging process. By constructing the actual condition map, it can intuitively reflect the difference between the state of the battery during the charging process and the standard state, and at the same time provide a basis for the next step of condition significance.

[0038] S300: Identify the similarity between the actual condition map and the standard condition map, and calculate to obtain the condition significance.

[0039] Furthermore, step S300 of the present application further includes:

[0040] S310: Collect a set of sample actual condition maps and a set of sample standard condition maps; S320: Use a siamese neural network, and use the set of sample actual condition maps and the set of sample standard condition maps to supervise and train the condition map similarity recognizer until the training converges; S330: Input the actual condition map and the standard condition map into the converged condition map similarity recognizer, and identify and output to obtain the similarity; S340: Subtract the similarity from 1 to obtain the condition significance.

[0041] Specifically, query the historical operation records of the same type of battery packs of the lithium battery pack, and collect the set of sample actual condition diagrams and the set of sample standard condition diagrams. Then, construct a condition diagram similarity recognizer based on the siamese neural network. The siamese neural network is a neural network architecture that includes two identical neural network branches, which share the same weights and receive two inputs. During training, the network compares the similarity of the two inputs and optimizes by calculating the difference between them. Then, use the set of sample actual condition diagrams and the set of sample standard condition diagrams as training data to perform supervised training on the condition diagram similarity recognizer. First, input the image pairs in the set of standard condition diagrams and the set of actual condition diagrams into the two branches of the siamese network, calculate the similarity value of the network output, and compare it with the true label (the similarity between the actual condition diagram and the standard condition diagram). The true similarity between the actual condition diagram and the standard condition diagram can be calculated as the ratio of the sum of the gray values of all pixel points of the actual condition diagram to the sum of the gray values of all pixel points of the standard condition diagram. Then, use the contrast loss function to calculate the loss value, and optimize the network weights through backpropagation, and continue training until the loss function converges to obtain the trained condition diagram similarity recognizer. During the training process, the siamese neural network will gradually learn how to judge the similarity by comparing the pixel values and features of the actual condition diagram and the standard condition diagram.

[0042] Then, input the actual condition diagram and the standard condition diagram into the converged condition diagram similarity recognizer for similarity recognition, and output the similarity between the two; and use 1 minus the similarity, and use the difference between the two as the condition significance. The condition significance reflects the gap between the working state and the ideal state of the battery pack during the charging process. The greater the condition significance, the greater the difference between the actual condition and the standard condition, which means the greater the probability of overcharging failure of the battery pack during the charging process and the higher the risk.

[0043] S400: According to the multiple voltage change significances, the multiple current change significances, and the multiple temperature change significances, perform overcharging failure prediction on the lithium battery pack to obtain a first overcharging failure coefficient, and use the condition significance to perform correction calculation on the first overcharging failure coefficient to obtain a second overcharging failure coefficient as the lithium battery failure prediction result.

[0044] Further, step S400 of the present application further includes:

[0045] S410: According to the overcharge fault detection data of lithium battery packs of the same model within a historical time, collect the set of sample voltage change significances, the set of sample current change significances, and the set of sample temperature change significances, and collect the proportions of overcharge faults occurring in the lithium battery packs under different sample voltage change significances, sample current change significances, and sample temperature change significances, to obtain the set of sample overcharge fault coefficients; S420: Use the set of sample voltage change significances, the set of sample current change significances, the set of sample temperature change significances, and the set of sample overcharge fault coefficients as supervised training data to train the overcharge fault predictor until the training converges; S430: Combine the multiple voltage change significances, the multiple current change significances, and the multiple temperature change significances to obtain multiple sets of input data, and respectively input them into the overcharge fault predictor to output and obtain multiple predicted overcharge fault coefficients; S440: Calculate the mean of the multiple predicted overcharge fault coefficients to obtain the first overcharge fault coefficient.

[0046] Specifically, according to the overcharge fault detection data of lithium battery packs of the same model within a historical time, collect the set of sample voltage change significances, the set of sample current change significances, and the set of sample temperature change significances; then count the proportions of overcharge faults occurring in the lithium battery packs under different sample voltage change significances, sample current change significances, and sample temperature change significances, that is, the ratio of the number of overcharge faults to the total number of charging times, and set the fault proportion as the sample overcharge fault coefficient to obtain the set of sample overcharge fault coefficients.

[0047] Construct an overcharge fault predictor based on a BP neural network. The overcharge fault predictor is a model that uses machine learning technology to predict overcharge faults during the charging process of a lithium battery pack, including an input layer, multiple hidden layers, and an output layer. Among them, the input data of the input layer are the voltage change significance, current change significance, and temperature change significance, and the output data of the output layer is the overcharge fault coefficient. Further, use the sample voltage change significance set, sample current change significance set, sample temperature change significance set, and sample overcharge fault coefficient set as supervised training data to perform supervised training on the overcharge fault predictor. First, transfer the inputs (voltage, current, temperature significance) in the training data to the input layer of the network. Then, each input value is multiplied by the weights of the input layer, added with a bias, and processed through an activation function to obtain the output of the hidden layer. The output of the hidden layer is then passed to the output layer, and through weighted summation and calculation by the activation function, the predicted overcharge fault coefficient is obtained. Then, use a loss function to calculate the difference between the network output and the true label (overcharge fault coefficient). Through the error of the output layer, backpropagate to calculate the errors of each hidden layer and the input layer. Through the gradient descent method, adjust the weights and biases of the network according to the errors of each layer. Through multiple iterations, the BP neural network gradually updates the weights and biases, making the predicted overcharge fault coefficient closer and closer to the true label until the loss function converges, and obtaining the trained overcharge fault predictor.

[0048] Next, combine the multiple voltage change significances, multiple current change significances, and multiple temperature change significances, that is, combine the change significances of each lithium battery cell into a set of input data to obtain multiple sets of input data. Then, input the multiple sets of input data into the overcharge fault predictor respectively, output multiple predicted overcharge fault coefficients, and calculate the mean of the multiple predicted overcharge fault coefficients, and set the mean calculation result as the first overcharge fault coefficient. Among them, the larger the overcharge fault coefficient, the greater the probability of the lithium battery pack having an overcharge fault and the higher the risk.

[0049] Further, step S400 of this application further includes:

[0050] S450: Calculate the sum of 1 and the conditional significance as the correction coefficient; S460: Multiply the first overcharge fault coefficient by the correction coefficient to obtain the second overcharge fault coefficient as the lithium battery fault prediction result.

[0051] Specifically, on the other hand, add 1 to the condition significance, and use the sum of the two as the correction coefficient. Then multiply the correction coefficient by the first overcharge fault coefficient, and use the product of the two as the second overcharge fault coefficient, and use the second overcharge fault coefficient as the lithium battery fault prediction result. By setting the correction coefficient according to the condition significance and correcting the first overcharge fault coefficient, the interference of the charging conditions during the charging process can be effectively considered, and the accuracy and reliability of the overcharge fault prediction can be further improved.

[0052] A lithium battery fault prediction method provided by an embodiment of the present invention has at least the following technical effects:

[0053] During the charging process of the lithium battery pack, collect the voltages, currents, and temperatures of multiple lithium battery cells in the lithium battery pack to form multiple voltage sequences, multiple current sequences, and multiple temperature sequences, and obtain the standard voltage sequence, standard current sequence, and standard temperature sequence of the lithium battery cell charging, and perform significant change analysis to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances; on the other hand, obtain the standard usage condition parameters of the lithium battery pack, construct a standard condition diagram, and collect the actual usage condition parameters of the current lithium battery pack to construct an actual condition diagram, wherein, perform lithium battery pack consistency analysis according to the multiple temperature change significances to obtain consistency condition parameters, and construct the actual condition diagram; then identify the similarity between the actual condition diagram and the standard condition diagram, and calculate to obtain the condition significance; further perform overcharge fault prediction on the lithium battery pack according to the multiple voltage change significances, multiple current change significances, and multiple temperature change significances to obtain the first overcharge fault coefficient; finally, perform correction calculation on the first overcharge fault coefficient using the condition significance to obtain the second overcharge fault coefficient as the lithium battery fault prediction result; that is to say, by fusing multi-dimensional data for lithium battery overcharge fault prediction, the accuracy and reliability of the overcharge fault prediction can be improved, so that potential risks during the charging process can be identified in advance, serious consequences caused by overcharge faults can be avoided, and the use safety of lithium batteries can be significantly improved.

[0054] Embodiment 2, as Figure 2As shown, based on the same inventive concept as the lithium battery fault prediction method provided in Embodiment 1, an embodiment of the present invention further provides a lithium battery fault prediction system, including: a change significance analysis module 01, configured to collect the voltages, currents, and temperatures of multiple lithium battery cells in a lithium battery pack during the charging process of the lithium battery pack, form multiple voltage sequences, multiple current sequences, and multiple temperature sequences, and obtain the standard voltage sequence, standard current sequence, and standard temperature sequence of the lithium battery cell charging, and perform change significance analysis to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances; an actual condition map construction module 02, configured to obtain the standard usage condition parameters of the lithium battery pack, construct a standard condition map, and collect the actual usage condition parameters of the current lithium battery pack, construct an actual condition map, wherein, according to the multiple temperature change significances, perform consistency analysis of the lithium battery pack to obtain consistency condition parameters, and construct the actual condition map; a condition significance calculation module 03, configured to identify the similarity between the actual condition map and the standard condition map, and calculate to obtain a condition significance; an overcharge fault prediction module 04, configured to perform overcharge fault prediction on the lithium battery pack according to the multiple voltage change significances, multiple current change significances, and multiple temperature change significances, obtain a first overcharge fault coefficient, and perform correction calculation on the first overcharge fault coefficient by using the condition significance to obtain a second overcharge fault coefficient as the lithium battery fault prediction result.

[0055] Further, the lithium battery fault prediction system further includes: during the charging process of the lithium battery pack, collect the voltages, currents, and temperatures of multiple lithium battery cells at multiple time nodes before the charge coefficient reaches the full charge threshold, form multiple voltage sequences, multiple current sequences, and multiple temperature sequences; obtain the standard voltage, standard current, and standard temperature at the multiple time nodes, and construct a standard voltage sequence, a standard current sequence, and a standard temperature sequence; calculate the deviations between the multiple voltage sequences, multiple current sequences, and multiple temperature sequences and the standard voltage sequence, standard current sequence, and standard temperature sequence respectively, to obtain multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences; according to the multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences, perform change significance analysis to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances.

[0056] Further, the lithium battery fault prediction system further includes: selecting a first selected voltage deviation within the first voltage deviation sequence, randomly selecting a first random voltage deviation, calculating the ratio of the first selected voltage deviation to the first random voltage deviation to obtain a first selected voltage change significance; continuing to calculate and obtain multiple first selected voltage change significances, and calculating the mean value to obtain a first voltage change significance; continuing to calculate and obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances according to the multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences.

[0057] Further, the lithium battery fault prediction system further includes: obtaining standard usage condition parameters of the lithium battery pack, where the standard usage condition parameters include a standard ambient temperature, a standard charging voltage, and a standard consistency parameter; setting the standard ambient temperature, the standard charging voltage, and the standard consistency parameter as preset values and using them as pixel values of image data to construct a standard condition map; collecting actual usage condition parameters of the current lithium battery pack, where the actual usage condition parameters include an actual ambient temperature, an actual charging voltage, and an actual consistency parameter, and the actual consistency parameter is obtained by calculating the variance of the multiple temperature change significances; calculating the ratios of the actual ambient temperature, the actual charging voltage, and the actual consistency parameter to the standard ambient temperature, the standard charging voltage, and the standard consistency parameter, multiplying by the preset value, and using them as pixel values of image data to construct an actual condition map.

[0058] Further, the lithium battery fault prediction system further includes: collecting a set of sample actual condition maps and a set of sample standard condition maps; using a siamese neural network to supervise and train a condition map similarity recognizer with the set of sample actual condition maps and the set of sample standard condition maps until the training converges; inputting the actual condition map and the standard condition map into the converged condition map similarity recognizer, and identifying and outputting a similarity; using 1 minus the similarity to obtain a condition significance.

[0059] Further, the lithium battery fault prediction system further includes: collecting a set of sample voltage change significances, a set of sample current change significances, and a set of sample temperature change significances based on overcharge fault detection data of lithium battery packs of the same model within a historical time, and collecting the proportions of overcharge faults occurring in the lithium battery packs under different sample voltage change significances, sample current change significances, and sample temperature change significances to obtain a set of sample overcharge fault coefficients; using the set of sample voltage change significances, the set of sample current change significances, the set of sample temperature change significances, and the set of sample overcharge fault coefficients as supervised training data to train an overcharge fault predictor until the training converges; combining the multiple voltage change significances, the multiple current change significances, and the multiple temperature change significances to obtain multiple sets of input data, respectively inputting the multiple sets of input data into the overcharge fault predictor, and outputting to obtain multiple predicted overcharge fault coefficients; calculating the mean of the multiple predicted overcharge fault coefficients to obtain a first overcharge fault coefficient.

[0060] Further, the lithium battery fault prediction system further includes: calculating the sum of 1 and the conditional significance as a correction coefficient; multiplying the first overcharge fault coefficient by the correction coefficient to obtain a second overcharge fault coefficient as the lithium battery fault prediction result.

[0061] Embodiment 3, please refer to Figure 3 , Figure 3 which is a schematic diagram of an embodiment of the electronic device provided in the embodiment of the present invention. As Figure 3As shown in the figure, an embodiment of the present invention provides an electronic device 500, including a memory 510, a processor 520, and a first computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the first computer program 511, the following steps are implemented: During the charging process of the lithium battery pack, collect the voltages, currents, and temperatures of multiple lithium battery cells in the lithium battery pack to form multiple voltage sequences, multiple current sequences, and multiple temperature sequences, and obtain the standard voltage sequence, standard current sequence, and standard temperature sequence of the lithium battery cell charging, and perform a significant change analysis to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances; obtain the standard usage condition parameters of the lithium battery pack, construct a standard condition graph, and collect the actual usage condition parameters of the current lithium battery pack to construct an actual condition graph, wherein, perform a consistency analysis of the lithium battery pack according to the multiple temperature change significances to obtain consistency condition parameters, and construct the actual condition graph; identify the similarity between the actual condition graph and the standard condition graph, and calculate to obtain a condition significance; according to the multiple voltage change significances, multiple current change significances, and multiple temperature change significances, perform an overcharge fault prediction on the lithium battery pack to obtain a first overcharge fault coefficient, and use the condition significance to perform a correction calculation on the first overcharge fault coefficient to obtain a second overcharge fault coefficient as the lithium battery fault prediction result.

[0062] Embodiment 4, please refer to Figure 4 , Figure 4 which is a schematic diagram of an embodiment of a computer-readable storage medium provided by an embodiment of the present invention. As Figure 4As shown in the figure, this embodiment provides a computer-readable storage medium 600, on which a second computer program 611 is stored. When the second computer program 611 is executed by a processor, the following steps are implemented: During the charging process of the lithium battery pack, collect the voltages, currents, and temperatures of multiple lithium battery cells in the lithium battery pack to form multiple voltage sequences, multiple current sequences, and multiple temperature sequences, and obtain the standard voltage sequence, standard current sequence, and standard temperature sequence of the lithium battery cell charging, and perform significant change analysis to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances; obtain the standard usage condition parameters of the lithium battery pack, construct a standard condition graph, and collect the actual usage condition parameters of the current lithium battery pack to construct an actual condition graph, where, according to the multiple temperature change significances, perform consistency analysis on the lithium battery pack to obtain consistency condition parameters and construct the actual condition graph; identify the similarity between the actual condition graph and the standard condition graph, and calculate to obtain a condition significance; according to the multiple voltage change significances, multiple current change significances, and multiple temperature change significances, perform overcharge fault prediction on the lithium battery pack to obtain a first overcharge fault coefficient, and use the condition significance to perform correction calculation on the first overcharge fault coefficient to obtain a second overcharge fault coefficient as the lithium battery fault prediction result.

[0063] It should be noted that in the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0064] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0065] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0066] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction means that implements the function specified in one or more of the processes and / or blocks Figure 1 one or more of the processes and / or blocks Figure 1 specified in one or more of the blocks or blocks.

[0067] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the function specified in one or more of the processes and / or blocks Figure 1 one or more of the processes and / or blocks Figure 1 specified in one or more of the blocks or blocks.

[0068] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept.

[0069] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A lithium battery failure prediction method, characterized in that: Methods include: During the charging process of the lithium battery pack, the voltage, current and temperature of multiple lithium battery cells in the lithium battery pack are collected to form multiple voltage sequences, multiple current sequences and multiple temperature sequences, and the standard voltage sequence, standard current sequence and standard temperature sequence for charging the lithium battery cells are obtained, and a change significance analysis is performed to obtain multiple voltage change significances, multiple current change significances and multiple temperature change significances; Obtain standard usage condition parameters of the lithium battery pack, construct a standard condition diagram, and collect actual usage condition parameters of the current lithium battery pack to construct an actual condition diagram, wherein the consistency analysis of the lithium battery pack is performed according to the multiple temperature change significances, the actual consistency parameters are obtained, and the actual condition diagram is constructed; Identify the similarity between the actual condition graph and the standard condition graph, and calculate the conditional saliency; According to the multiple voltage change significances, the multiple current change significances, and the multiple temperature change significances, an overcharge fault prediction of the lithium battery pack is performed to obtain a first overcharge fault coefficient, and the first overcharge fault coefficient is corrected and calculated using the conditional significance to obtain a second overcharge fault coefficient as a lithium battery fault prediction result; The standard use condition parameters of the lithium battery pack are obtained to construct a standard condition diagram, and the actual use condition parameters of the current lithium battery pack are collected to construct an actual condition diagram, including: Obtaining standard usage condition parameters of a lithium battery pack, wherein the standard usage condition parameters include standard ambient temperature, standard charging voltage, and standard consistency parameters; The standard ambient temperature, standard charging voltage, and standard consistency parameters are set as preset values ​​as pixel values ​​of image data to construct a standard condition map; Collecting actual use condition parameters of the current lithium battery pack, wherein the actual use condition parameters include actual ambient temperature, actual charging voltage and actual consistency parameters, and the actual consistency parameters are obtained by calculating the variance of the plurality of temperature change significances; The ratio of the actual ambient temperature, the actual charging voltage and the actual consistency parameter to the standard ambient temperature, the standard charging voltage and the standard consistency parameter is calculated and multiplied by the preset value as the pixel value of the image data to construct an actual condition map.

2. A lithium battery failure prediction method according to claim 1, characterized in that: The voltage, current and temperature of multiple lithium battery cells in the lithium battery pack are collected to form multiple voltage sequences, multiple current sequences and multiple temperature sequences, and the standard voltage sequence, standard current sequence and standard temperature sequence for charging the lithium battery cells are obtained, and the change significance analysis is performed to obtain multiple voltage change significances, multiple current change significances and multiple temperature change significances, including: During the charging process of the lithium battery pack, the voltage, current and temperature of multiple lithium battery cells at multiple time points before the charge factor reaches the full threshold are collected to form multiple voltage sequences, multiple current sequences and multiple temperature sequences; Acquire the standard voltage, standard current and standard temperature in the multiple time nodes, and construct a standard voltage sequence, a standard current sequence and a standard temperature sequence; Respectively calculating deviations between the multiple voltage sequences, the multiple current sequences, and the multiple temperature sequences and the standard voltage sequence, the standard current sequence, and the standard temperature sequence to obtain multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences; A change significance analysis is performed according to the multiple voltage deviation sequences, multiple current deviation sequences, and multiple temperature deviation sequences to obtain multiple voltage change significances, multiple current change significances, and multiple temperature change significances.

3. A lithium battery fault prediction method according to claim 2, characterized in that: According to the multiple voltage deviation sequences, multiple current deviation sequences and multiple temperature deviation sequences, a change significance analysis is performed to obtain multiple voltage change significances, multiple current change significances and multiple temperature change significances, including: Selecting a first selected voltage deviation in a first voltage deviation sequence, and randomly selecting a first random voltage deviation, calculating a ratio of the first selected voltage deviation to the first random voltage deviation, and obtaining a first selected voltage change significance; Continue to calculate to obtain multiple first selected voltage change significances, and calculate the average to obtain the first voltage change significance; Continuously calculate and obtain a plurality of voltage change significances, a plurality of current change significances, and a plurality of temperature change significances according to the plurality of voltage deviation sequences, the plurality of current deviation sequences, and the plurality of temperature deviation sequences.

4. A lithium battery failure prediction method according to claim 1, characterized in that: Identifying the similarity between the actual condition graph and the standard condition graph and calculating the conditional saliency includes: Collecting a set of sample actual condition graphs and a set of sample standard condition graphs; Using a twin neural network, using the sample actual condition graph set and the sample standard condition graph set, supervise the training of a condition graph similarity identifier until the training converges; Inputting the actual condition graph and the standard condition graph into the converged condition graph similarity identifier, and identifying the output to obtain similarity; The conditional saliency is obtained by subtracting the similarity from 1.

5. A lithium battery failure prediction method according to claim 1, characterized in that: According to the multiple voltage change significances, the multiple current change significances, and the multiple temperature change significances, an overcharge fault prediction of the lithium battery pack is performed to obtain a first overcharge fault coefficient, including: According to the overcharge fault detection data of the same type of lithium battery pack in the historical time, the sample voltage change significance set, the sample current change significance set and the sample temperature change significance set are collected, and the proportion of overcharge faults of the lithium battery pack under different sample voltage change significances, sample current change significances and sample temperature change significances are collected to obtain the sample overcharge fault coefficient set; Using the sample voltage change significance set, the sample current change significance set, the sample temperature change significance set and the sample overcharge fault coefficient set as supervised training data, training an overcharge fault predictor until the training converges; Combining the multiple voltage change significances, the multiple current change significances and the multiple temperature change significances to obtain multiple groups of input data, respectively inputting them into the overcharge fault predictor, and outputting multiple predicted overcharge fault coefficients; An average value of the plurality of predicted overcharge fault coefficients is calculated to obtain a first overcharge fault coefficient.

6. A lithium battery fault prediction method according to claim 1, characterized in that: The first overcharge fault coefficient is corrected and calculated by using the condition significance to obtain a second overcharge fault coefficient, including: Calculate the sum of 1 and the conditional significance as the correction coefficient; The correction coefficient is multiplied by the first overcharge fault coefficient to obtain a second overcharge fault coefficient as a lithium battery fault prediction result.

7. A lithium battery failure prediction system, characterized in that: The steps for implementing a lithium battery fault prediction method according to any one of claims 1 to 6 include: A change significance analysis module is used to collect the voltage, current and temperature of multiple lithium battery cells in the lithium battery pack during the charging process of the lithium battery pack, form multiple voltage sequences, multiple current sequences and multiple temperature sequences, and obtain the standard voltage sequence, standard current sequence and standard temperature sequence for charging the lithium battery cells, and perform change significance analysis to obtain multiple voltage change significances, multiple current change significances and multiple temperature change significances; An actual condition map construction module is used to obtain standard use condition parameters of the lithium battery pack, construct a standard condition map, and collect actual use condition parameters of the current lithium battery pack to construct an actual condition map, wherein the consistency analysis of the lithium battery pack is performed according to the multiple temperature change significances, the actual consistency parameters are obtained, and the actual condition map is constructed; A conditional saliency calculation module, used to identify the similarity between the actual condition graph and the standard condition graph, and calculate the conditional saliency; An overcharge fault prediction module is used to predict the overcharge fault of the lithium battery pack according to the multiple voltage change significances, the multiple current change significances and the multiple temperature change significances, obtain a first overcharge fault coefficient, use the conditional significance to perform correction calculation on the first overcharge fault coefficient, and obtain a second overcharge fault coefficient as a lithium battery fault prediction result.

8. An electronic device, characterized in that: include: Memory for storing computer software programs; A processor is used to read and execute the computer software program, thereby implementing the steps of a lithium battery fault prediction method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium, characterized in that: The storage medium stores a computer software program, and when the computer software program is executed by the processor, the steps of a lithium battery fault prediction method as described in any one of claims 1 to 6 are implemented.

Images