Lithium ion battery health state evaluation method and apparatus, and electronic device

By constructing a state evaluation model of one-dimensional convolutional residual network and feature sharing network, combining the fully connected network, outputting the SOH value and irreversible expansion of lithium-ion batteries, the problem of high dependence on historical data in the prior art is solved. A more accurate assessment of the health status of lithium-ion batteries is achieved.

CN118938056APending Publication Date: 2024-11-12GUANGDONG POWER GRID CO LTD +1
View PDF 0 Cites 8 Cited by

Patent Information

Application Number
CN202410992577.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-23
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

In the prior art, when performing the status evaluation of lithium-ion battery based on data-driven methods, the dependence on the full amount of data in the historical time period is high, resulting in low evaluation accuracy.

Method used

The state evaluation model is constructed using one-dimensional convolutional residual network, feature sharing network and fully connected network. By collecting current sequence, voltage sequence and temperature sequence, the SOH value and irreversible expansion of lithium-ion batteries are output, and the preset threshold is used for comparison. Yes, determine the battery health status.

Benefits of technology

Reduce the dependence on the full data of the lithium-ion battery life cycle, improve the evaluation accuracy, and obtain accurate evaluation results by intercepting a small piece of data for status evaluation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118938056A_ABST
    Figure CN118938056A_ABST
Patent Text Reader

Abstract

The invention discloses a lithium ion battery health state evaluation method and device and electronic equipment, and relates to the technical field of batteries, and the method comprises the steps: collecting a current sequence, a voltage sequence and a temperature sequence of a lithium ion battery in a target time period, and obtaining a capacity sequence; inputting the current sequence, the voltage sequence, the temperature sequence and the capacity sequence into a state evaluation model, and outputting the SOH value and the irreversible swelling amount of the lithium ion battery; comparing the SOH value of the lithium ion battery with a preset SOH threshold value to obtain a first comparison result; comparing the irreversible swelling capacity of the lithium ion battery with a preset irreversible swelling capacity threshold value to obtain a second comparison result; and determining the health state of the battery based on the first comparison result and the second comparison result. According to the method and the device, the technical problems of relatively high dependence on total data in a historical time period and relatively low evaluation accuracy of state evaluation of the lithium ion battery based on a data driving method in related technologies are solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of lithium-ion batteries, and in particular to a method for evaluating the health status of lithium-ion batteries, a device thereof, and electronic equipment. Background Art

[0002] At present, the lithium-ion battery industry is developing rapidly and is widely used in fields such as grid energy storage, electric vehicles, and aerospace. The thickness change of lithium-ion batteries due to their expansion behavior during the cycle is one of the main reasons for capacity reduction and safety problems. Accurate estimation of the SOH value and irreversible expansion information of lithium-ion batteries can provide more comprehensive battery health assessment information and is an important guarantee for the safe and stable operation of the system. According to the joint estimation results of the SOH and irreversible expansion of lithium-ion batteries, it can be used to guide battery replacement when the SOH drops to a certain level and the irreversible expansion reaches the warning value to ensure the stable operation of the equipment; in addition, in the battery system design stage, the battery system structure design can also be optimized based on the SOH value and irreversible expansion of lithium-ion batteries to improve the overall performance of the battery system.

[0003] In the related technology, the lithium-ion battery state estimation method can be based on a data-driven method. Based on the data-driven state estimation method, feature extraction is the key to accuracy and robustness. Feature extraction is generally divided into two types: manual feature extraction based on domain knowledge and automatic feature extraction based on deep learning models. However, both of the above methods are highly dependent on the full amount of data in the historical time period, which leads to low accuracy in the health status assessment of lithium-ion batteries.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a method for assessing the health status of a lithium-ion battery, a device thereof, and an electronic device, so as to at least solve the technical problem in the related art that the health status assessment of a lithium-ion battery based on a data-driven method has a high dependence on the full amount of data in a historical time period and a low assessment accuracy.

[0006] According to one aspect of an embodiment of the present invention, a method for assessing the health status of a lithium-ion battery is provided, comprising: collecting a current sequence, a voltage sequence, and a temperature sequence of a lithium-ion battery within a target time period, and obtaining a capacity sequence of the lithium-ion battery based on the current sequence and a sampling frequency; inputting the current sequence, the voltage sequence, the temperature sequence, and the capacity sequence into a state assessment model, and outputting a SOH value and an irreversible expansion amount of the lithium-ion battery based on the state assessment model, wherein the state assessment model is a pre-trained model for jointly predicting the SOH value and the irreversible expansion amount of the lithium-ion battery, and the state assessment model is a pre-trained model for jointly predicting the SOH value and the irreversible expansion amount of the lithium-ion battery. The estimation model includes: a one-dimensional convolutional residual network, a feature sharing network and a fully connected network, the SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount; the SOH value of the lithium-ion battery is compared with a preset SOH threshold to obtain a first comparison result; the irreversible expansion amount of the lithium-ion battery is compared with a preset irreversible expansion amount threshold to obtain a second comparison result; and the health state of the lithium-ion battery is determined based on the first comparison result and the second comparison result.

[0007] Optionally, the step of inputting the current sequence, voltage sequence, temperature sequence and capacity sequence into a state assessment model, and outputting the SOH value and irreversible expansion amount of the lithium-ion battery based on the state assessment model includes: inputting the current sequence, voltage sequence, temperature sequence and capacity sequence into a one-dimensional convolutional residual network of the state assessment model, and outputting a feature vector of the lithium-ion battery through the one-dimensional convolutional residual network; inputting the feature vector into a feature sharing network of the state assessment model, and outputting a shared feature vector through the feature sharing network; inputting the shared feature vector into a fully connected network of the state assessment model, and outputting the SOH value and irreversible expansion amount of the lithium-ion battery through the fully connected network.

[0008] Optionally, the step of determining the health status of the lithium-ion battery based on the first comparison result and the second comparison result includes: determining that the lithium-ion battery is in an unhealthy state when at least one of the first comparison result and the second comparison result hits a state evaluation rule, wherein the state evaluation rule includes: the SOH value of the lithium-ion battery is less than a preset SOH threshold, and the irreversible expansion amount of the lithium-ion battery is greater than a preset irreversible expansion amount threshold; and generating an alarm message for the lithium-ion battery in an unhealthy state.

[0009] Optionally, the step of training the state assessment model includes: performing an aging cycle test on the lithium-ion battery, and constructing sample data based on the aging cycle test, wherein the aging cycle test represents cyclic constant current charging and constant current discharging of the lithium-ion battery; dividing the sample data into a training set and a test set in turn based on a cross-validation method to obtain K groups of training sets and test sets, wherein K is a positive integer; constructing an initial state assessment model; training and testing the initial state assessment model based on the K groups of training sets and test sets to obtain K state assessment models; and integrating the K state assessment models to obtain the final state assessment model.

[0010] Optionally, the step of performing the aging cycle test on the lithium-ion battery and constructing sample data based on the aging cycle test includes: selecting M unused lithium-ion batteries of the same model, and performing N groups of aging cycle tests on each of the lithium-ion batteries, where M and N are both positive integers; collecting voltage data, current data, temperature data, SOH value data and irreversible expansion data of the lithium-ion battery during each group of aging cycle tests; constructing a voltage sequence, current sequence, temperature sequence and capacity sequence of the lithium-ion battery in each group of aging cycle tests based on the voltage data, the current data and the temperature data; constructing a state feature of the lithium-ion battery based on the voltage sequence, current sequence, temperature sequence and capacity sequence; and obtaining the sample data based on the voltage sequence, current sequence, temperature sequence and capacity sequence, SOH value data, irreversible expansion data and state features of the lithium-ion battery.

[0011] Optionally, the step of constructing a voltage sequence, a current sequence, a temperature sequence and a capacity sequence of the lithium-ion battery in each group of aging cycle tests based on the voltage data, the current data and the temperature data includes: for each group of the aging cycle tests, drawing a voltage-current curve and a voltage-temperature curve based on the voltage data, the current data and the temperature data; selecting a voltage interval so that the voltage-current curve and the voltage-temperature curve include the voltage interval, and sampling the voltage data, the current data and the temperature data in the voltage interval to obtain the voltage sequence, the current sequence and the temperature sequence of the lithium-ion battery in the aging cycle test; obtaining a sampling frequency, and calculating the capacity sequence of the lithium-ion battery in the aging cycle test based on the sampling frequency and the current sequence.

[0012] Optionally, the initial state assessment model is trained and tested based on the K groups of training sets and test sets to obtain K state assessment models, including the following steps: step one, selecting a group of training sets and test sets as target training sets and target test sets, iteratively training the initial state assessment model based on the target training sets to obtain the trained state assessment model, and testing the trained state assessment model based on the target test set to obtain a test result, and when the test result indicates that the trained state assessment model passes the test, a trained state assessment model is obtained; repeating step one, each time selecting different target training sets and target test sets to construct a state assessment model, until K groups of cyclic training are completed to obtain K trained state assessment models.

[0013] According to another aspect of an embodiment of the present invention, a health status assessment device for a lithium-ion battery is provided, comprising: a collection unit, for collecting a current sequence, a voltage sequence and a temperature sequence of the lithium-ion battery within a target time period, and obtaining a capacity sequence of the lithium-ion battery based on the current sequence and a sampling frequency; an output unit, for inputting the current sequence, the voltage sequence, the temperature sequence and the capacity sequence into a state assessment model, and outputting the SOH value and the irreversible expansion amount of the lithium-ion battery based on the state assessment model, wherein the state assessment model is a pre-trained model for jointly predicting the SOH value and the irreversible expansion amount of the lithium-ion battery, and the state assessment model comprises : a one-dimensional convolutional residual network, a feature sharing network and a fully connected network, the SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount; a first comparison unit is used to compare the SOH value of the lithium-ion battery with a preset SOH threshold to obtain a first comparison result; a second comparison unit is used to compare the irreversible expansion amount of the lithium-ion battery with a preset irreversible expansion amount threshold to obtain a second comparison result; a determination unit is used to determine the health state of the lithium-ion battery based on the first comparison result and the second comparison result.

[0014] Optionally, the output unit includes: a first output subunit, used to input the current sequence, voltage sequence, temperature sequence and capacity sequence into a one-dimensional convolutional residual network of the state assessment model, and output a feature vector of the lithium-ion battery through the one-dimensional convolutional residual network; a second output subunit, used to input the feature vector into a feature sharing network of the state assessment model, and output a shared feature vector through the feature sharing network; and a third output subunit, used to input the shared feature vector into a fully connected network of the state assessment model, and output the SOH value and irreversible expansion amount of the lithium-ion battery through the fully connected network.

[0015] Optionally, the determination unit includes: a first determination subunit, used to determine that the lithium-ion battery is in an unhealthy state when at least one of the first comparison result and the second comparison result hits a state evaluation rule, wherein the state evaluation rule includes: the SOH value of the lithium-ion battery is less than a preset SOH threshold, and the irreversible expansion amount of the lithium-ion battery is greater than a preset irreversible expansion amount threshold; a first generation subunit, used to generate alarm information for the lithium-ion battery in an unhealthy state.

[0016] Optionally, the health status assessment device of the lithium-ion battery also includes: a first testing subunit, used to perform an aging cycle test on the lithium-ion battery, and construct sample data based on the aging cycle test, wherein the aging cycle test represents cyclic constant current charging and constant current discharging of the lithium-ion battery; a first division subunit, used to divide the sample data into a training set and a test set in sequence based on a cross-validation method to obtain K groups of training sets and test sets, wherein K is a positive integer; a first construction subunit, used to construct an initial state assessment model; a first training subunit, used to train and test the initial state assessment model based on the K groups of training sets and test sets to obtain K state assessment models; a first integration subunit, used to integrate the K state assessment models to obtain the final state assessment model.

[0017] Optionally, the first test subunit includes: a first selection module, used to select M unused lithium-ion batteries of the same model, and perform N groups of aging cycle tests on each of the lithium-ion batteries, where M and N are both positive integers; a first acquisition module, used to collect voltage data, current data, temperature data, SOH value data and irreversible expansion data of the lithium-ion battery during each group of aging cycle tests; a first construction module, used to construct a voltage sequence, current sequence, temperature sequence and capacity sequence of the lithium-ion battery in each group of aging cycle tests based on the voltage data, the current data and the temperature data; a second construction module, used to construct a state feature of the lithium-ion battery based on the voltage sequence, current sequence, temperature sequence and capacity sequence; a first acquisition module, used to obtain the sample data based on the voltage sequence, current sequence, temperature sequence and capacity sequence, SOH value data, irreversible expansion data and state features of the lithium-ion battery.

[0018] Optionally, the first construction module includes: a first drawing submodule, used to draw a voltage-current curve and a voltage-temperature curve based on the voltage data, the current data and the temperature data for each group of the aging cycle test; a first sampling submodule, used to select a voltage interval so that the voltage-current curve and the voltage-temperature curve contain the voltage interval, and sample the voltage data, the current data and the temperature data in the voltage interval to obtain the voltage sequence, the current sequence and the temperature sequence of the lithium-ion battery in the aging cycle test; a first calculation submodule, used to obtain a sampling frequency, and calculate the capacity sequence of the lithium-ion battery in the aging cycle test based on the sampling frequency and the current sequence.

[0019] Optionally, the first training subunit includes: a first training module, used for step one, selecting a group of training sets and test sets as target training sets and target test sets, iteratively training the initial state assessment model based on the target training set to obtain the trained state assessment model, and testing the trained state assessment model based on the target test set to obtain a test result, and when the test result indicates that the trained state assessment model passes the test, obtaining the trained state assessment model; a first repetition module, used to repeat step one, each time selecting different target training sets and target test sets to construct the state assessment model, until K groups of cyclic training are completed to obtain K trained state assessment models.

[0020] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned lithium-ion battery health status assessment methods.

[0021] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the above-mentioned methods for assessing the health status of a lithium-ion battery.

[0022] In the present application, the following steps are performed: first, the current sequence, voltage sequence and temperature sequence of the lithium-ion battery within the target time period are collected, and the capacity sequence of the lithium-ion battery is obtained based on the current sequence and the sampling frequency, and then the current sequence, voltage sequence, temperature sequence and capacity sequence are input into a state assessment model, and the SOH value and irreversible expansion amount of the lithium-ion battery are output based on the state assessment model, and then the SOH value of the lithium-ion battery is compared with a preset SOH threshold to obtain a first comparison result, and the irreversible expansion amount of the lithium-ion battery is compared with a preset irreversible expansion amount threshold to obtain a second comparison result; finally, the health state of the lithium-ion battery is determined based on the first comparison result and the second comparison result.

[0023] In the present application, when evaluating the health status of a lithium-ion battery, data such as current series, voltage series, temperature series, and capacity series are collected over a period of time. Through a pre-constructed joint model, the SOH value and irreversible expansion amount of the lithium-ion battery can be jointly estimated, so that the health status of the lithium-ion battery can be accurately evaluated based on the SOH value and the irreversible expansion amount. The dependence on the full data of the lithium-ion battery life cycle is weak, and the status evaluation can be performed by intercepting a small piece of data to obtain accurate evaluation results, thereby solving the technical problem in the related art that the status evaluation of lithium-ion batteries based on data-driven methods is highly dependent on the full data in the historical time period and the evaluation accuracy is low. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0025] Figure 1 is a flow chart of an optional method for assessing the health status of a lithium-ion battery according to an embodiment of the present invention;

[0026] Figure 2 is a schematic diagram of an optional lithium-ion battery health status assessment process according to an embodiment of the present invention;

[0027] Figure 3 is a schematic diagram of an optional lithium-ion battery health status assessment device according to an embodiment of the present invention;

[0028] Figure 4 The present invention is a hardware structure block diagram of an electronic device (or mobile device) for evaluating the health status of a lithium-ion battery according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] To facilitate those skilled in the art to understand the present invention, some terms or nouns involved in various embodiments of the present invention are explained below:

[0032] CBAM module, an attention mechanism module used in deep learning, is called Convolutional BlockAttention Module. It is mainly used to improve the correlation between different feature maps of the network, thereby improving the performance of the model.

[0033] It should be noted that the health status assessment method and device of lithium-ion batteries in the present application can be used in the field of battery technology when the encrypted data is verified based on the blockchain, and can also be used in any field outside the field of blockchain technology when the encrypted data is verified based on the blockchain. The application of the health status assessment method and device of lithium-ion batteries in the present application does not limit the application field.

[0034] The following embodiments of the present invention can be applied to various lithium-ion battery health status assessment systems / applications / devices. The present invention collects segment voltage, segment current and segment temperature data during constant current charging, and uses the current sequence, temperature sequence, statistical characteristics of the capacity sequence in the voltage-capacity curve and the voltage-capacity curve in the segment voltage interval as input variables of the model to train the state assessment model, thereby achieving a joint estimation of the SOH value and the irreversible expansion amount, and improving the accuracy of the state estimation.

[0035] The present invention constructs a joint estimation model of battery SOH and irreversible expansion based on a one-dimensional convolutional residual network, a feature sharing network and a fully connected network that integrates channel attention mechanism and spatial attention mechanism modules, thereby obtaining a state evaluation model that can perform state evaluation, and can accurately estimate the SOH value and irreversible expansion of lithium-ion batteries in different usage scenarios. This evaluation method takes into account the influence of temperature and rate on the charging characteristics of lithium-ion batteries during the charging process, and does not require the use of complete constant current data of the charging stage. The use conditions are relatively simple, and it is suitable for SOH and irreversible expansion estimation and state evaluation of lithium-ion batteries in a wide temperature range and at a large charging rate.

[0036] The present invention is described in detail below in conjunction with various embodiments.

[0037] Embodiment 1

[0038] According to an embodiment of the present invention, an embodiment of a method for assessing the health status of a lithium-ion battery is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0039] Figure 1 is a flow chart of an optional method for evaluating the health status of a lithium-ion battery according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0040] Step S101, collecting a current sequence, a voltage sequence and a temperature sequence of a lithium-ion battery within a target time period, and obtaining a capacity sequence of the lithium-ion battery based on the current sequence and the sampling frequency;

[0041] Step S102, inputting the current sequence, voltage sequence, temperature sequence and capacity sequence into a state assessment model, and outputting the SOH value and irreversible expansion amount of the lithium-ion battery based on the state assessment model, wherein the state assessment model is a pre-trained model for jointly predicting the SOH value and irreversible expansion amount of the lithium-ion battery, and the state assessment model includes: a one-dimensional convolutional residual network, a feature sharing network and a fully connected network, the SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount;

[0042] Step S103, comparing the SOH value of the lithium-ion battery with a preset SOH threshold to obtain a first comparison result;

[0043] Step S104, comparing the irreversible expansion amount of the lithium-ion battery with a preset irreversible expansion amount threshold to obtain a second comparison result;

[0044] Step S105 , determining the health status of the lithium-ion battery based on the first comparison result and the second comparison result.

[0045] Through the above steps, firstly, the current sequence, voltage sequence and temperature sequence of the lithium-ion battery in the target time period are collected, and the capacity sequence of the lithium-ion battery is obtained based on the current sequence and the sampling frequency, and then the current sequence, voltage sequence, temperature sequence and capacity sequence are input into the state assessment model, and the SOH value and irreversible expansion amount of the lithium-ion battery are output based on the state assessment model, and then the SOH value of the lithium-ion battery is compared with a preset SOH threshold to obtain a first comparison result, and the irreversible expansion amount of the lithium-ion battery is compared with a preset irreversible expansion amount threshold to obtain a second comparison result; finally, the health state of the lithium-ion battery is determined based on the first comparison result and the second comparison result.

[0046] In this embodiment, when evaluating the health status of the lithium-ion battery, data such as the current sequence, voltage sequence, temperature sequence, and capacity sequence within a period of time are collected. Through a pre-constructed joint model, the SOH value and the irreversible expansion amount of the lithium-ion battery can be jointly estimated, so that the health status of the lithium-ion battery can be accurately evaluated based on the SOH value and the irreversible expansion amount. The dependence on the full data of the lithium-ion battery life cycle is weak, and the status evaluation can be performed by intercepting a small piece of data to obtain accurate evaluation results, thereby solving the technical problem in the related art that the status evaluation of lithium-ion batteries based on the data-driven method is highly dependent on the full data in the historical time period and the evaluation accuracy is low.

[0047] The embodiment of the present invention is described in detail below in combination with the above steps.

[0048] As an indispensable energy storage unit in modern electronic devices and electric vehicles, the performance and life of lithium-ion batteries have always been hot issues in research and application. During use, lithium-ion batteries will experience different degrees of loss, which not only affects the capacity and energy output of the battery, but may also cause safety problems. Therefore, it is particularly important to evaluate and predict its health status. Through a comprehensive analysis of basic parameters such as battery capacity, number of charge and discharge cycles, and expansion behavior, the current health status of lithium-ion batteries can be evaluated and their future performance changes can be predicted. This evaluation not only helps to extend the service life of the battery, but also ensures its safety and reliability in various application scenarios.

[0049] It should be noted that the SOH value (State of Health) and irreversible expansion volume of lithium-ion batteries are important indicators for evaluating the performance and life of lithium-ion batteries. The SOH value can characterize the capacity change of lithium-ion batteries during use, and the battery capacity is an important indicator for comprehensively evaluating the aging status of batteries. The irreversible expansion volume refers to the permanent increase in battery volume due to gas generated inside the lithium battery or other reasons. When the surface of the lithium-ion battery expands to a certain extent, it will cause safety hazards to the lithium-ion battery, and the lithium-ion battery with hidden dangers needs to be safely handled.

[0050] Step S101 , collecting a current sequence, a voltage sequence and a temperature sequence of a lithium-ion battery within a target time period, and obtaining a capacity sequence of the lithium-ion battery based on the current sequence and the sampling frequency.

[0051] It should be noted that the embodiments of the present invention can predict the SOH value and irreversible expansion amount of the lithium-ion battery in a future time period based on the segment current data, segment voltage data, segment temperature data during the battery charging and discharging process and the battery capacity data calculated based on the current data, and then evaluate the health status of the lithium-ion battery.

[0052] It should be noted that the current data, voltage data and temperature data of the lithium-ion battery in the target time period are collected, and each segment of the data is sampled based on a preset sampling frequency to obtain the current sequence, voltage sequence and temperature sequence of the lithium-ion battery in the target time period, and then the capacity sequence of the lithium-ion battery is calculated based on the current value and sampling frequency of the current sequence, thereby obtaining the input data of the state assessment model to predict and evaluate the health status of the lithium-ion battery.

[0053] Step S102, inputting the current sequence, voltage sequence, temperature sequence and capacity sequence into a state assessment model, and outputting the SOH value and irreversible expansion amount of the lithium-ion battery based on the state assessment model.

[0054] In this embodiment, the state assessment model is a pre-trained model for jointly predicting the SOH value and irreversible expansion amount of a lithium-ion battery. The state assessment model includes: a one-dimensional convolutional residual network, a feature sharing network, and a fully connected network. The SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount.

[0055] It should be noted that after obtaining the current sequence, voltage sequence, temperature sequence and capacity sequence of the lithium-ion battery, they are input into a pre-trained state assessment model. The state assessment model can extract the state features of the lithium-ion battery and jointly calculate the SOH value and the irreversible expansion amount based on a multi-layer network structure, thereby outputting the estimated SOH value and irreversible expansion amount of the lithium-ion battery in the future, which are used to evaluate the health status of the lithium-ion battery and estimate the degree of aging of the lithium-ion battery.

[0056] Optionally, the step of training the state assessment model includes: performing an aging cycle test on the lithium-ion battery, and constructing sample data based on the aging cycle test, wherein the aging cycle test represents cyclically performing constant current charging and constant current discharging on the lithium-ion battery; dividing the sample data into a training set and a test set in turn based on a cross-validation method to obtain K groups of training sets and test sets, wherein K is a positive integer; constructing an initial state assessment model; training and testing the initial state assessment model based on the K groups of training sets and test sets to obtain K state assessment models; and integrating the K state assessment models to obtain a final state assessment model.

[0057] The state assessment model of the embodiment of the present application can jointly predict the SOH value and the non-expansion amount of the lithium-ion battery. When constructing and training the lithium-ion battery, first, an aging cycle test is performed on multiple unused lithium-ion batteries of the same type at a varying temperature to simulate the charging and discharging process of the lithium-ion battery, and the terminal voltage, current, temperature, SOH value and irreversible expansion amount in the thickness direction of the lithium-ion battery during each cycle of the lithium-ion battery are collected and used as training data for model training. When training the model, the training data is divided into multiple subsets by cross-validation, and the subsets are used as training sets and test sets. The initial state assessment model is iteratively trained and tested to obtain multiple state assessment models, and the multiple state assessment models are integrated to obtain a trained state assessment model.

[0058] It should be noted that the state assessment model of the embodiment of the present invention consists of three parts: a one-dimensional convolutional residual network, a feature sharing network and a fully connected network. The one-dimensional convolutional residual network is used to extract the state characteristics of the lithium-ion battery. The feature sharing network is used to share feature information in multiple learning tasks to improve the accuracy of the prediction results. The fully connected network is used to predict the SOH value and irreversible expansion of the lithium-ion battery based on the state characteristics.

[0059] When constructing the initial state assessment model, first build a one-dimensional convolutional residual network fused with the CBAM module, set the activation function type, number of layers, filter size, sliding step size and padding size of the one-dimensional convolutional residual network model fused with the CBAM module, and then construct the one-dimensional convolutional residual network model F fused with the CBAM module. CABM-CNN (x) is expressed as: o1 = F CABM-CNN (x), where o1 represents the output variable of the one-dimensional convolutional residual network model fused with the CBAM module.

[0060] Then, a feature sharing network based on a deep neural network is constructed, and the number of layers, activation function type, and number of neurons of the deep neural network model are set. DNN (x) is expressed as: share =F DNN (x), where X share Represents the feature sharing network output variable.

[0061] Finally, a fully connected network of the lithium-ion battery is constructed. Based on the fully connected network, the SOH value and irreversible expansion of the lithium-ion battery are calculated. The expression of the fully connected network is as follows: in, represents the SOH value predicted by the model, represents the irreversible expansion predicted by the model.

[0062] Finally, according to the one-dimensional convolutional residual network, feature sharing network and fully connected network integrated with CBAM, a joint estimation model of SOH and irreversible expansion of lithium-ion batteries based on feature fusion is constructed.

[0063] Optionally, the steps of performing an aging cycle test on a lithium-ion battery and constructing sample data based on the aging cycle test include: selecting M unused lithium-ion batteries of the same model, performing N groups of aging cycle tests on each lithium-ion battery, where M and N are both positive integers; collecting voltage data, current data, temperature data, SOH value data and irreversible expansion data of the lithium-ion battery during each group of aging cycle tests; constructing a voltage sequence, current sequence, temperature sequence and capacity sequence of the lithium-ion battery in each group of aging cycle tests based on the voltage data, current data and temperature data; constructing a state feature of the lithium-ion battery based on the voltage sequence, current sequence, temperature sequence and capacity sequence; and obtaining sample data based on the voltage sequence, current sequence, temperature sequence and capacity sequence, SOH value data, irreversible expansion data and state features of the lithium-ion battery.

[0064] It should be noted that, when constructing sample data for model training, the embodiment of the present invention first selects M unused lithium-ion batteries of the same model, and performs multiple groups of aging cycle tests on each lithium-ion battery. For each group of aging cycle tests, the voltage data, current data, temperature data, SOH value of the lithium-ion battery in this aging cycle, and irreversible expansion amount in the thickness direction of the lithium-ion battery are collected, and then the above voltage data, current data, and temperature data are sampled to obtain current sequence, voltage sequence, temperature sequence and other data, and state features are extracted from the sequence data. The sequence data, state features, SOH value, and irreversible expansion amount in the thickness direction of the lithium-ion battery of N groups of cycle aging tests of M lithium-ion batteries of the same type are used as sample data to cross-train the state assessment model.

[0065] Optionally, the step of constructing a voltage sequence, current sequence, temperature sequence and capacity sequence of the lithium-ion battery in each group of aging cycle tests based on the voltage data, current data and temperature data includes: for each group of aging cycle tests, drawing a voltage-current curve and a voltage-temperature curve based on the voltage data, current data and temperature data; selecting a voltage interval so that the voltage-current curve and the voltage-temperature curve include the voltage interval, and sampling the voltage data, current data and temperature data in the voltage interval to obtain the voltage sequence, current sequence and temperature sequence of the lithium-ion battery in the aging cycle test; obtaining a sampling frequency, and calculating the capacity sequence of the lithium-ion battery in the aging cycle test based on the sampling frequency and the current sequence.

[0066] It should be noted that when obtaining sequence data such as voltage sequence, current sequence, temperature sequence and capacity sequence, firstly, the voltage-current curve and voltage-temperature curve are drawn according to the voltage data, current data and temperature data of each group of aging cycle tests, and then a fixed voltage interval is selected so that both the voltage-current curve and the voltage-temperature curve include the voltage interval, and then the voltage, current and temperature data within the voltage interval are sampled to obtain the voltage sequence, current sequence and temperature sequence data, and the capacity sequence of the lithium-ion battery in the voltage interval is calculated according to the sampling frequency and the current value of the current sequence.

[0067] Optionally, the initial state assessment model is trained and tested based on K groups of training sets and test sets to obtain K state assessment models, including the following steps: step one, selecting a group of training sets and test sets as target training sets and target test sets, iteratively training the initial state assessment model based on the target training sets to obtain a trained state assessment model, and testing the trained state assessment model based on the target test set to obtain a test result, and when the test result indicates that the trained state assessment model passes the test, a trained state assessment model is obtained; repeating step one, selecting different target training sets and target test sets each time to construct the state assessment model, until K groups of cyclic training are completed to obtain K trained state assessment models.

[0068] It should be noted that after obtaining the sample data, the state assessment model is trained based on cross-validation. The cross-validation divides the sample data into K subsets, K-1 subsets of which are used as training sets, and the remaining subset is used as a test set. The initial state assessment model is iteratively trained, and different training sets and test sets are selected in each cycle until all subsets are used as test sets for model training, thereby obtaining K state assessment models, and finally integrating the K models to obtain the final state assessment model. The integration mechanism may include integration based on vector machines, or integration based on random forests, etc. The embodiment of the present invention does not limit the integration mechanism of multiple models.

[0069] Optionally, the current sequence, voltage sequence, temperature sequence and capacity sequence are input into the state assessment model, and the step of outputting the SOH value and irreversible expansion amount of the lithium-ion battery based on the state assessment model includes: inputting the current sequence, voltage sequence, temperature sequence and capacity sequence into the one-dimensional convolutional residual network of the state assessment model, and outputting the feature vector of the lithium-ion battery through the one-dimensional convolutional residual network; inputting the feature vector into the feature sharing network of the state assessment model, and outputting the shared feature vector through the feature sharing network; inputting the shared feature vector into the fully connected network of the state assessment model, and outputting the SOH value and irreversible expansion amount of the lithium-ion battery through the fully connected network.

[0070] It should be noted that when estimating the SOH value and irreversible expansion of lithium-ion batteries, the current sequence, temperature sequence, voltage sequence and capacity sequence are input into the state assessment model in turn. The state assessment model is based on the type of lithium-ion batteries. First, the feature extraction is automatically performed through a one-dimensional convolutional residual network to obtain the feature vector of the lithium-ion battery. The feature vector is then input into the feature sharing network of the state assessment model. The feature sharing network can obtain shared feature information from multiple historical tasks, and then apply it to this prediction process, and screen the feature vectors, select the feature vectors that are important to the prediction results, and obtain the shared feature vectors. Finally, the shared feature vectors are input into the fully connected network of the state assessment model, and the SOH value and irreversible expansion of the lithium-ion battery are estimated based on the fully connected network, and output as the output result.

[0071] Step S103, comparing the SOH value of the lithium-ion battery with a preset SOH threshold to obtain a first comparison result.

[0072] Step S104, comparing the irreversible expansion amount of the lithium-ion battery with a preset irreversible expansion amount threshold to obtain a second comparison result.

[0073] The SOH value and irreversible expansion amount in the future period are obtained through the output of the state assessment model, and compared with the preset thresholds. The health status of the lithium-ion battery can be known based on the comparison results of the SOH value and the irreversible expansion amount with the thresholds, or the service life of the lithium-ion battery can be determined based on the comparison results.

[0074] Step S105 , determining the health status of the lithium-ion battery based on the first comparison result and the second comparison result.

[0075] Optionally, the step of determining the health status of the lithium-ion battery based on the first comparison result and the second comparison result includes: determining that the lithium-ion battery is in an unhealthy state when at least one of the first comparison result and the second comparison result hits a state evaluation rule, wherein the state evaluation rule includes: the SOH value of the lithium-ion battery is less than a preset SOH threshold, and the irreversible expansion amount of the lithium-ion battery is greater than a preset irreversible expansion amount threshold; and generating an alarm message for the lithium-ion battery in an unhealthy state.

[0076] It should be noted that, for lithium-ion batteries, the SOH value is the ratio between the battery capacity calculated in real time and the nominal battery capacity of the battery when it leaves the factory. Therefore, the larger the SOH value, the better the battery health status. If the SOH value output by the model is less than the preset SOH threshold, it indicates that the health status of the lithium-ion battery is poor and the lithium-ion battery needs to be replaced or other processing operations. The irreversible expansion amount is the ratio between the expansion amount of the lithium-ion battery in the thickness direction and the initial expansion amount. The irreversible expansion amount is a phenomenon formed during the use process with the chemical reaction during the battery charging and discharging process. When the irreversible expansion amount increases to a certain extent, the battery is at risk and needs to be replaced. Therefore, the irreversible expansion amount output by the model can be used to evaluate the service life and danger level of the lithium-ion battery. If any one of the SOH value and the irreversible expansion amount of the lithium-ion battery exceeds the preset threshold, it indicates that the lithium-ion battery is in an unhealthy state, and an alarm message is generated for the lithium-ion battery in the unhealthy state to prompt relevant personnel to pay attention to the battery dynamics and replace the battery if necessary.

[0077] Another optional specific implementation is described in detail below.

[0078] Figure 2 is a schematic diagram of an optional lithium-ion battery health status assessment process according to an embodiment of the present invention, such as Figure 2 As shown in Figure 1, the health status assessment process of lithium-ion batteries includes:

[0079] Step 1: Data measurement of lithium-ion batteries;

[0080] Perform aging cycle tests on multiple unused lithium batteries of the same model. The aging cycle tests are performed by charging at a constant current and constant voltage, and then fully discharging at a constant current after being fully charged.

[0081] The battery data such as the terminal voltage, current, temperature, SOH value and irreversible expansion in the thickness direction of each lithium-ion battery are collected. Among them, the ratio of the current maximum discharge capacity to the nominal capacity is defined as SOH, and the difference between the current minimum expansion and the initial expansion is defined as irreversible expansion.

[0082] From the battery data obtained, the current, voltage, temperature, health status and irreversible expansion of each lithium-ion battery in a constant current charging process are intercepted, and the N-cycle cycle data of the s-th battery is The specific form is as follows:

[0083]

[0084] Where, ns ,i Indicates the data length of the sth battery in the i-th cycle; Represents the current data of the sth battery in the i-th cycle; Represents the voltage data of the sth battery in the i-th circle; Represents the temperature data of the sth battery in the i-th cycle; Indicates the health status of the sth battery after i cycles; It represents the irreversible expansion of the sth battery after undergoing i cycles. Based on this, the single cell data of each lithium-ion battery is obtained.

[0085] Step 2: determine a fixed voltage interval and obtain the voltage, current and temperature sequence corresponding to the interval;

[0086] Based on the single-cell data of each lithium-ion battery, a change curve of the internal capacity and irreversible expansion of the lithium-ion battery during its entire life, and a change curve of the voltage, current and temperature during constant current charging are constructed;

[0087] Select a fixed voltage range [V strat ,V end ], so that all charging voltage, current and temperature curves include this interval, and the sth battery falls in the voltage interval [V strat ,V end ] on the voltage sequence Current sequence and temperature series

[0088]

[0089] In the formula, Indicates that the sth battery falls in the voltage range [V strat ,V end ] on the data length; They represent that the sth battery falls in the voltage range [V strat ,V end ]The sampling values ​​of current, voltage and temperature at the t+i-th data point on ].

[0090] Step three, feature extraction;

[0091] Step 4, obtaining the current mean, temperature mean, mean, standard deviation, skewness of the capacity sequence in the voltage-capacity curve, and the capacity sequence in the voltage-capacity curve to obtain sample data;

[0092] Select a fixed voltage range [V strat ,V end ], with a voltage interval of ΔV, a set of voltage sequences V inter , the expression of voltage series is as follows:

[0093]

[0094] In the formula, Indicates rounding down, V start+i Represents the voltage sequence V inter The voltage value of the i-th position in , M represents the number of interpolation points in the fixed voltage interval.

[0095] With voltage as the independent variable and charge capacity as the dependent variable, for two data points and By linear interpolation, we can get a point V between these two points. start+i The corresponding capacity value As shown below:

[0096]

[0097] According to the current sequence and temperature series And the capacity series in the voltage-capacity curve Calculate the current mean, temperature mean, and mean of the capacity sequence in the voltage-capacity curve. Each mean is shown in the following formula:

[0098]

[0099] In the formula, and They represent that the sth battery falls in the voltage range [V strat ,V end ] The average current and temperature on the Indicates that the sth battery falls in the voltage range [V strat ,V end ] is the mean value of the capacity series in the voltage-capacity curve.

[0100] Calculate the standard deviation and skewness of the capacity sequence in the voltage-capacity curve. The expressions of standard deviation and skewness are as follows:

[0101]

[0102] In the formula, and They represent that the sth battery falls in the voltage range [V strat ,V end ]The standard deviation and skewness of the capacity series in the voltage-capacity curve on .

[0103] The current mean, temperature mean, mean, standard deviation and skewness of the capacity sequence in the voltage-capacity curve and the capacity sequence in the voltage-capacity curve are taken as features related to the battery health state and irreversible expansion, and are used as inputs of the joint estimation model of the battery health state and irreversible expansion, as shown below:

[0104]

[0105] In the formula, It represents the input variables of the model when estimating the battery health status and irreversible expansion of the sth battery after the kth cycle process, including two parts, Represents the pre-extracted state features, including: current mean, temperature mean, and mean, standard deviation, and skewness of the capacity sequence in the voltage-capacity curve, etc. Represents the capacity sequence in the voltage-capacity curve of the s-th battery after the k-th cycle process.

[0106] According to the current sequence Combined with the sampling time interval Δt, the charging capacity sequence in this interval As shown in the following formula;

[0107]

[0108] Sample data for model training is obtained based on single-cell data of lithium-ion batteries and feature data extracted from the single-cell data.

[0109] Step 5: Divide the sample data into training set and test set;

[0110] The sample data is divided into K subsets. When training the model, a cross-validation model is adopted for training. In each cycle of iterative training, K-1 subsets among the K subsets are used as training sets, and the other subset is used as a test set. For each cycle, a different subset is selected as a test set, so that the sample data can be fully utilized to build a more accurate state assessment model.

[0111] Step 6: construct a joint estimation model of SOH and irreversible expansion of lithium-ion batteries (i.e., the above-mentioned state assessment model);

[0112] Construct a one-dimensional convolutional residual network fused with the CBAM module, set the activation function type, number of layers, filter size, sliding step size and padding size of the one-dimensional convolutional residual network model fused with the CBAM module, and construct the one-dimensional convolutional residual network model F fused with the CBAM module. CABM-CNN (x) is represented as follows:

[0113] o1=F CABM-CNN (x)

[0114] Where o1 represents the output variable of the one-dimensional convolutional residual network model fused with the CBAM module.

[0115] Construct a feature sharing network based on a deep neural network, set the number of deep neural network model layers, activation function type and number of neurons, and construct a feature sharing network based on a deep neural network F. DNN (x) is represented as follows:

[0116] X share =F DNN (x)

[0117] Where, X share Represents the feature sharing network output variable.

[0118] Based on the fully connected network, the SOH and irreversible expansion of the lithium-ion battery are calculated as follows:

[0119]

[0120] Where, X share Represents the feature sharing network output variable.

[0121] According to the one-dimensional convolutional residual network, feature sharing network and fully connected network integrated with CBAM, a joint estimation model of SOH and irreversible expansion of lithium-ion batteries based on feature fusion is constructed. Combined with the input features of the model, the SOH and irreversible expansion results of the s-th battery after the k-th cycle are estimated by the model. As shown below:

[0122]

[0123] In the formula, [·] indicates the concatenation and fusion of two feature variables.

[0124] Using the feature training set D after feature extraction train , the joint estimation model of SOH and irreversible expansion of lithium-ion batteries is trained, and the target loss is the mean square error loss:

[0125]

[0126] Where Y i is the target value; is the estimated value of the model; N is the number of training samples; y i,so h and y i,i et are the true values ​​of SOH and irreversible expansion, respectively; and are the estimated values ​​of SOH and irreversible expansion, respectively.

[0127] The joint estimation model of SOH and irreversible expansion of lithium-ion batteries is tested according to the test set to achieve estimation of SOH and irreversible expansion of lithium-ion batteries. If the test passes, a trained state assessment model is obtained.

[0128] Step 7, iteratively train the model to obtain a trained model;

[0129] Step 8: Test the model based on the test set;

[0130] Step nine, obtaining the trained state evaluation model;

[0131] Step 10: Conduct status assessment on the lithium-ion battery.

[0132] The battery health status (i.e., SOH value) and irreversible expansion as the output results of the state assessment model can be expressed as: represents the output data of the state assessment model, represents the SOH value output by the state assessment model, Represents the irreversible expansion output of the state assessment model.

[0133] In an embodiment of the present invention, the joint estimation of the SOH and irreversible expansion of lithium-ion batteries can be achieved only through fragmentary data during the constant current charging process of some batteries. At the same time, a one-dimensional convolutional residual network based on the fusion of CBAM is constructed to realize the adaptive extraction of data-driven features in different usage scenarios of lithium-ion batteries; a feature sharing network is constructed to realize the fusion of statistical features and data-driven features of capacity sequences in current, temperature, and voltage-capacity curves; and the estimation of SOH and irreversible expansion of lithium-ion batteries is realized based on a fully connected network. This method can, to a certain extent, solve the problem that the state of lithium-ion batteries is difficult to accurately estimate due to the difficulty in obtaining complete constant current charging measurement data, the lack of sufficient fusion of artificial features and data-driven features, and the difficulty in adaptively extracting data-driven features. It is suitable for the estimation of SOH and irreversible expansion of lithium-ion batteries under a wide temperature range and a large charging rate.

[0134] The following is a detailed description in conjunction with another embodiment.

[0135] Embodiment 2

[0136] A lithium-ion battery health status assessment device provided in this embodiment includes multiple implementation units, each implementation unit corresponds to each implementation step in the above-mentioned embodiment 1. Its specific implementation methods and beneficial effects can refer to the above-mentioned method embodiments and will not be repeated here.

[0137] Figure 3is a schematic diagram of an optional lithium-ion battery health status assessment device according to an embodiment of the present invention, such as Figure 3 As shown, the health status assessment device of the lithium-ion battery may include: a collection unit 31, an output unit 32, a first comparison unit 33, a second comparison unit 34, and a determination unit 35, wherein:

[0138] The acquisition unit 31 is used to acquire the current sequence, voltage sequence and temperature sequence of the lithium-ion battery within the target time period, and obtain the capacity sequence of the lithium-ion battery based on the current sequence and the sampling frequency;

[0139] An output unit 32 is used to input the current sequence, voltage sequence, temperature sequence and capacity sequence into a state assessment model, and output the SOH value and irreversible expansion amount of the lithium-ion battery based on the state assessment model, wherein the state assessment model is a pre-trained model for jointly predicting the SOH value and irreversible expansion amount of the lithium-ion battery, and the state assessment model includes: a one-dimensional convolutional residual network, a feature sharing network and a fully connected network, the SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount;

[0140] A first comparison unit 33, used to compare the SOH value of the lithium-ion battery with a preset SOH threshold value to obtain a first comparison result;

[0141] A second comparison unit 34, used to compare the irreversible expansion amount of the lithium-ion battery with a preset irreversible expansion amount threshold to obtain a second comparison result;

[0142] The determination unit 35 is used to determine the health state of the lithium-ion battery based on the first comparison result and the second comparison result.

[0143] The above-mentioned lithium-ion battery health status assessment device collects the current sequence, voltage sequence and temperature sequence of the lithium-ion battery in the target time period through the acquisition unit 31, and obtains the capacity sequence of the lithium-ion battery based on the current sequence and the sampling frequency; the current sequence, voltage sequence, temperature sequence and capacity sequence are input into the state assessment model through the output unit 32, and the SOH value and irreversible expansion amount of the lithium-ion battery are output based on the state assessment model, wherein the state assessment model is a pre-trained model for jointly predicting the SOH value and irreversible expansion amount of the lithium-ion battery, and the state assessment model includes: a one-dimensional convolutional residual network , feature sharing network and fully connected network, the SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount; the SOH value of the lithium-ion battery is compared with a preset SOH threshold value by a first comparison unit 33 to obtain a first comparison result; the irreversible expansion amount of the lithium-ion battery is compared with a preset irreversible expansion amount threshold value by a second comparison unit 34 to obtain a second comparison result; the health status of the lithium-ion battery is determined by a determination unit 35 based on the first comparison result and the second comparison result.

[0144] In this embodiment, when evaluating the health status of the lithium-ion battery, data such as the current sequence, voltage sequence, temperature sequence, and capacity sequence within a period of time are collected. Through a pre-constructed joint model, the SOH value and the irreversible expansion amount of the lithium-ion battery can be jointly estimated, so that the health status of the lithium-ion battery can be accurately evaluated based on the SOH value and the irreversible expansion amount. The dependence on the full data of the lithium-ion battery life cycle is weak, and the status evaluation can be performed by intercepting a small piece of data to obtain accurate evaluation results, thereby solving the technical problem in the related art that the status evaluation of lithium-ion batteries based on the data-driven method is highly dependent on the full data in the historical time period and the evaluation accuracy is low.

[0145] Optionally, the output unit 32 includes: a first output subunit, used to input the current sequence, voltage sequence, temperature sequence and capacity sequence into the one-dimensional convolution residual network of the state assessment model, and output the feature vector of the lithium-ion battery through the one-dimensional convolution residual network; a second output subunit, used to input the feature vector into the feature sharing network of the state assessment model, and output the shared feature vector through the feature sharing network; a third output subunit, used to input the shared feature vector into the fully connected network of the state assessment model, and output the SOH value and irreversible expansion amount of the lithium-ion battery through the fully connected network.

[0146] Optionally, the determination unit 35 includes: a first determination subunit, used to determine that the lithium-ion battery is in an unhealthy state when at least one of the first comparison result and the second comparison result hits a state evaluation rule, wherein the state evaluation rule includes: the SOH value of the lithium-ion battery is less than a preset SOH threshold, and the irreversible expansion amount of the lithium-ion battery is greater than a preset irreversible expansion amount threshold; a first generation subunit, used to generate alarm information for the lithium-ion battery in an unhealthy state.

[0147] Optionally, the health status assessment device of a lithium-ion battery also includes: a first testing subunit, used to perform an aging cycle test on the lithium-ion battery, and construct sample data based on the aging cycle test, wherein the aging cycle test represents constant current charging and constant current discharging of the lithium-ion battery; a first division subunit, used to divide the sample data into a training set and a test set in sequence based on a cross-validation method, to obtain K groups of training sets and test sets, wherein K is a positive integer; a first construction subunit, used to construct an initial status assessment model; a first training subunit, used to train and test the initial status assessment model based on K groups of training sets and test sets, to obtain K status assessment models; a first integration subunit, used to integrate the K status assessment models to obtain a final status assessment model.

[0148] Optionally, the first test subunit includes: a first selection module, used to select M unused lithium-ion batteries of the same model, and perform N groups of aging cycle tests on each lithium-ion battery, where M and N are both positive integers; a first acquisition module, used to collect voltage data, current data, temperature data, SOH value data and irreversible expansion data of the lithium-ion battery during each group of aging cycle tests; a first construction module, used to construct a voltage sequence, current sequence, temperature sequence and capacity sequence of the lithium-ion battery in each group of aging cycle tests based on the voltage data, current data and temperature data; a second construction module, used to construct a state feature of the lithium-ion battery based on the voltage sequence, current sequence, temperature sequence and capacity sequence; a first acquisition module, used to obtain sample data based on the voltage sequence, current sequence, temperature sequence and capacity sequence, SOH value data, irreversible expansion data and state features of the lithium-ion battery.

[0149] Optionally, the first construction module includes: a first drawing submodule, which is used to draw a voltage-current curve and a voltage-temperature curve based on the voltage data, current data and temperature data for each group of aging cycle tests; a first sampling submodule, which is used to select a voltage interval so that the voltage-current curve and the voltage-temperature curve contain the voltage interval, and sample the voltage data, current data and temperature data in the voltage interval to obtain a voltage sequence, a current sequence and a temperature sequence of the lithium-ion battery in the aging cycle test; and a first calculation submodule, which is used to obtain a sampling frequency, and calculate the capacity sequence of the lithium-ion battery in the aging cycle test based on the sampling frequency and the current sequence.

[0150] Optionally, the first training subunit includes: a first training module, used for step one, selecting a group of training sets and test sets as target training sets and target test sets, iteratively training the initial state assessment model based on the target training set to obtain a trained state assessment model, and testing the trained state assessment model based on the target test set to obtain a test result, and when the test result indicates that the trained state assessment model passes the test, a trained state assessment model is obtained; a first repetition module, used to repeat step one, each time selecting different target training sets and target test sets to construct a state assessment model, until K groups of cyclic training are completed to obtain K trained state assessment models.

[0151] The above-mentioned lithium-ion battery health status assessment device may also include a processor and a memory. The above-mentioned acquisition unit 31, output unit 32, first comparison unit 33, second comparison unit 34, determination unit 35, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0152] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the health status of the lithium-ion battery can be evaluated by adjusting kernel parameters.

[0153] The above-mentioned memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0154] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, the computer-readable storage medium including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned lithium-ion battery health status assessment methods.

[0155] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by one or more processors, the one or more processors implement any one of the above-mentioned methods for assessing the health status of a lithium-ion battery.

[0156] According to another aspect of an embodiment of the present invention, a computer program product is further provided. The computer program product includes a computer program, wherein when the computer program is executed by a processor, any of the above-mentioned methods for assessing the health status of a lithium-ion battery is implemented.

[0157] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with the following method steps: collecting a current sequence, a voltage sequence, and a temperature sequence of a lithium-ion battery within a target time period, and obtaining a capacity sequence of the lithium-ion battery based on the current sequence and the sampling frequency; inputting the current sequence, the voltage sequence, the temperature sequence, and the capacity sequence into a state assessment model, and outputting a SOH value and an irreversible expansion amount of the lithium-ion battery based on the state assessment model, wherein the state assessment model is pre-trained and used for jointly predicting the SOH value and the irreversible expansion amount of the lithium-ion battery. The state assessment model includes: a one-dimensional convolutional residual network, a feature sharing network and a fully connected network, the SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount; the SOH value of the lithium-ion battery is compared with a preset SOH threshold to obtain a first comparison result; the irreversible expansion amount of the lithium-ion battery is compared with a preset irreversible expansion amount threshold to obtain a second comparison result; and the health state of the lithium-ion battery is determined based on the first comparison result and the second comparison result.

[0158] Figure 4 1 is a hardware structure block diagram of an electronic device (or mobile device) according to a method for evaluating the health status of a lithium-ion battery in accordance with an embodiment of the present invention. Figure 4 As shown, the electronic device may include one or more ( Figure 4 402a, 402b, ..., 402n are used to illustrate) processor 402 (processor 402 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), memory 404 for storing data. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a keyboard, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 4The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 4 More or fewer components as shown, or with Figure 4 Different configurations are shown.

[0159] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0160] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0161] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0162] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0163] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0164] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0165] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for evaluating the health status of a lithium-ion battery, It is characterized in that include: Collecting a current sequence, a voltage sequence, and a temperature sequence of a lithium-ion battery within a target time period, and obtaining a capacity sequence of the lithium-ion battery based on the current sequence and a sampling frequency; The current sequence, the voltage sequence, the temperature sequence and the capacity sequence are input into a state assessment model, and the SOH value and the irreversible expansion amount of the lithium-ion battery are output based on the state assessment model, wherein the state assessment model is a pre-trained model for jointly predicting the SOH value and the irreversible expansion amount of the lithium-ion battery, and the state assessment model includes: a one-dimensional convolutional residual network, a feature sharing network and a fully connected network, the SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount; Comparing the SOH value of the lithium-ion battery with a preset SOH threshold to obtain a first comparison result; Comparing the irreversible expansion amount of the lithium-ion battery with a preset irreversible expansion amount threshold to obtain a second comparison result; The health state of the lithium-ion battery is determined based on the first comparison result and the second comparison result.

2. The method according to claim 1, It is characterized in that The step of inputting the current sequence, the voltage sequence, the temperature sequence and the capacity sequence into a state assessment model, and outputting the SOH value and the irreversible expansion amount of the lithium-ion battery based on the state assessment model comprises: Inputting the current sequence, the voltage sequence, the temperature sequence and the capacity sequence into a one-dimensional convolutional residual network of the state assessment model, and outputting a feature vector of the lithium-ion battery through the one-dimensional convolutional residual network; Inputting the feature vector into a feature sharing network of the state assessment model, and outputting a shared feature vector through the feature sharing network; The shared feature vector is input into a fully connected network of the state assessment model, and the SOH value and irreversible expansion amount of the lithium-ion battery are output through the fully connected network.

3. The method according to claim 1, It is characterized in that The step of determining the health status of the lithium-ion battery based on the first comparison result and the second comparison result comprises: In the case where at least one of the first comparison result and the second comparison result hits a state evaluation rule, determining that the lithium ion battery is in an unhealthy state, wherein the state evaluation rule includes: the SOH value of the lithium ion battery is less than a preset SOH threshold, and the irreversible expansion amount of the lithium ion battery is greater than a preset irreversible expansion amount threshold; A warning message is generated for the lithium-ion battery that is in an unhealthy state.

4. The method according to claim 1, It is characterized in that The steps of training the state assessment model include: Performing an aging cycle test on the lithium-ion battery, and constructing sample data based on the aging cycle test, wherein the aging cycle test represents cyclically performing constant current charging and constant current discharging on the lithium-ion battery; Based on the cross-validation method, the sample data is divided into a training set and a test set in turn to obtain K groups of training sets and test sets, where K is a positive integer; Constructing an initial state assessment model; Training and testing the initial state assessment model based on the K groups of training sets and test sets to obtain K state assessment models; The K state assessment models are integrated to obtain the final state assessment model.

5. The method according to claim 4, It is characterized in that The steps of performing the aging cycle test on the lithium-ion battery and constructing sample data based on the aging cycle test include: Select M unused lithium-ion batteries of the same model, and perform N groups of aging cycle tests on each of the lithium-ion batteries, where M and N are both positive integers; Collecting voltage data, current data, temperature data, SOH value data and irreversible expansion data of the lithium-ion battery during each aging cycle test; Constructing a voltage sequence, a current sequence, a temperature sequence and a capacity sequence of the lithium-ion battery in each group of aging cycle tests based on the voltage data, the current data and the temperature data; Constructing the state characteristics of the lithium-ion battery based on the voltage sequence, current sequence, temperature sequence and capacity sequence; The sample data is obtained based on the voltage sequence, current sequence, temperature sequence and capacity sequence, SOH value data, irreversible expansion data and state characteristics of the lithium-ion battery.

6. The method according to claim 5, It is characterized in that The step of constructing a voltage sequence, a current sequence, a temperature sequence and a capacity sequence of the lithium-ion battery in each group of aging cycle tests based on the voltage data, the current data and the temperature data comprises: For each group of the aging cycle tests, a voltage-current curve and a voltage-temperature curve are plotted based on the voltage data, the current data and the temperature data; Selecting a voltage interval so that the voltage-current curve and the voltage-temperature curve include the voltage interval, and sampling the voltage data, the current data and the temperature data in the voltage interval to obtain the voltage sequence, the current sequence and the temperature sequence of the lithium-ion battery in the aging cycle test; A sampling frequency is obtained, and a capacity sequence of the lithium-ion battery in the aging cycle test is calculated based on the sampling frequency and the current sequence.

7. The method according to claim 4, It is characterized in that The steps of training and testing the initial state assessment model based on the K groups of training sets and test sets to obtain K state assessment models include: Step 1, selecting a set of training sets and test sets as target training sets and target test sets, iteratively training the initial state assessment model based on the target training set to obtain the trained state assessment model, and testing the trained state assessment model based on the target test set to obtain a test result, and when the test result indicates that the trained state assessment model passes the test, obtaining the trained state assessment model; Repeat step 1, each time selecting a different target training set and a target test set to construct a state assessment model, until K groups of cyclic training are completed, and K trained state assessment models are obtained.

8. A device for evaluating the health status of a lithium-ion battery, It is characterized in that include: A collection unit, used to collect the current sequence, voltage sequence and temperature sequence of the lithium-ion battery within a target time period, and obtain the capacity sequence of the lithium-ion battery based on the current sequence and the sampling frequency; an output unit, used to input the current sequence, the voltage sequence, the temperature sequence and the capacity sequence into a state assessment model, and output the SOH value and the irreversible expansion amount of the lithium-ion battery based on the state assessment model, wherein the state assessment model is a pre-trained model for jointly predicting the SOH value and the irreversible expansion amount of the lithium-ion battery, and the state assessment model includes: a one-dimensional convolutional residual network, a feature sharing network and a fully connected network, the SOH value represents the ratio of the current battery capacity of the lithium-ion battery to the initial battery capacity, and the irreversible expansion amount represents the ratio of the current minimum expansion amount of the lithium-ion battery in the thickness direction to the initial expansion amount; A first comparison unit, used to compare the SOH value of the lithium-ion battery with a preset SOH threshold value to obtain a first comparison result; a second comparing unit, used to compare the irreversible expansion amount of the lithium-ion battery with a preset irreversible expansion amount threshold to obtain a second comparing result; A determination unit is used to determine the health state of the lithium-ion battery based on the first comparison result and the second comparison result.

9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the health status assessment method of a lithium-ion battery according to any one of claims 1 to 7.

10. An electronic device, It is characterized in that It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the health status assessment method of a lithium-ion battery as described in any one of claims 1 to 7.

Citation Information

Cited By

  • Lithium battery health state assessment method based on deep learning

    CN119535238A

  • Retired battery health state evaluation method and system based on auto-encoder

    CN119535239A

  • Battery capacity analysis and prediction method and system of storage battery

    CN119619849A

  • A method and system for analyzing and predicting the battery capacity of a storage battery

    CN119619849B

  • Lithium battery health state rapid assessment method and system

    CN119916225A