A lithium battery capacity detection method and system

Abstract

The invention discloses a lithium battery capacity detection method, comprising the following steps: performing a charge and discharge test on a sample lithium battery to obtain sample test capacity information, and obtaining sample ultrasonic feedback information of the sample lithium battery through ultrasonic detection technology; training a first convolutional neural network according to the sample ultrasonic feedback information; performing a charge and discharge test on a lithium battery to be tested to obtain test capacity information and test ultrasonic feedback information, and using the first convolutional neural network to obtain ultrasonic capacity information; obtaining the capacity of the lithium battery to be tested according to the test capacity information and ultrasonic capacity information; evaluating the health state of the lithium battery to be tested, and obtaining the battery fault by using a fault diagnosis model; obtaining the capacity of the lithium battery to be tested based on the test capacity and ultrasonic capacity, thereby improving the accuracy of the lithium battery capacity detection, ensuring the accuracy of the battery capacity detection, and obtaining the battery fault of the lithium battery to be tested by using a fault model, which is conducive to timely handling of the battery fault.

Description

Technical Field

[0001] The present invention relates to the technical field of measuring electrical variables, and in particular to a lithium battery capacity detection method and system. Background Art

[0002] Due to the vigorous research and development of new energy vehicles, its main energy supply part, lithium-ion power batteries, have also been extensively studied. Accurately measuring the battery capacity of lithium-ion power batteries is of great significance to the safe driving of new energy vehicles and timely replenishment of power.

[0003] At present, the Chinese invention patent with application number 202310744741.X provides a ternary lithium battery capacity detection method based on machine learning. Although the battery capacity detection model is trained using a data set and the newly collected battery data is input into the battery capacity detection model, real-time online detection of the battery capacity can be achieved. However, using a single model to obtain the battery capacity cannot guarantee the accuracy of the battery capacity and is prone to errors. In addition, there is no fault detection for lithium batteries with unqualified battery capacity, which is not conducive to ensuring battery safety and reducing battery technical failures. Summary of the invention

[0004] The technical problem solved by the present invention is that the prior art uses a single model to obtain battery capacity, which cannot guarantee the accuracy of the battery capacity and is prone to errors. In addition, there is no fault detection for lithium batteries with unqualified battery capacity, which is not conducive to ensuring battery safety and reducing battery technical failures.

[0005] In order to solve the above technical problems, in a first aspect, the present invention provides a lithium battery capacity detection method, comprising the following steps:

[0006] Step S1, performing charge and discharge tests on sample lithium batteries respectively, obtaining sample test capacity information, and obtaining sample ultrasonic feedback information of each sample lithium battery respectively through ultrasonic detection technology;

[0007] Step S2, training a first convolutional neural network according to the sample ultrasound feedback information to obtain a first convolutional neural network model;

[0008] Step S3, performing a charge and discharge test on the lithium battery to be tested, obtaining test capacitance information and ultrasonic feedback information to be tested, and obtaining ultrasonic capacitance information by using the first convolutional neural network through the ultrasonic feedback information to be tested;

[0009] Step S4, obtaining the capacity of the lithium battery to be tested according to the test capacity information and the ultrasonic capacity information;

[0010] Step S5, evaluating the health status of the lithium battery to be tested according to the electric capacity, and obtaining the battery fault of the lithium battery by using a fault diagnosis model;

[0011] As a preferred embodiment of the lithium battery capacity detection method described in the present invention, wherein:

[0012] The step S1 specifically includes the following steps:

[0013] Step S101, randomly selecting n sample lithium batteries that have the same specifications as the lithium battery to be tested, where n is a natural number;

[0014] Step S102, at a constant temperature T, use a constant current to perform a charge and discharge test on n sample lithium batteries respectively, charge the sample lithium batteries with a constant current until the charge cut-off voltage is reached, then discharge the sample lithium batteries with a constant current until the discharge cut-off voltage is reached, and monitor the voltage of the sample lithium batteries in real time, and record the discharge time used to reach the cut-off voltage , the cut-off voltage includes a charge cut-off voltage and a discharge cut-off voltage;

[0015] Step S103, obtaining sample test capacity information of each sample lithium battery according to the constant current, discharge time and cut-off voltage, wherein the sample test capacity information includes the sample test capacity;

[0016] Step S104, applying an ultrasonic pulse signal to the lithium battery by using an ultrasonic transmitter, the ultrasonic pulse signal is sent to the ultrasonic receiving device after passing through the sample lithium battery, the ultrasonic receiving device transmits the acquired sample ultrasonic feedback information to the signal processing device, the sample ultrasonic feedback information includes charging ultrasonic feedback information and discharging ultrasonic feedback information, the sample ultrasonic feedback information specifically includes the internal ultrasonic image of the sample lithium battery and ultrasonic feedback wave information, the ultrasonic feedback wave information includes the waveform, amplitude and peak-to-peak value of the feedback ultrasonic wave;

[0017] As a preferred embodiment of the lithium battery capacity detection method described in the present invention, wherein:

[0018] The step S2 specifically includes the following steps:

[0019] Step S201, the signal processing device receives the sample ultrasonic feedback information, and processes the feedback ultrasonic wave by using a wavelet transform method to obtain a time-frequency image of the sample feedback ultrasonic wave;

[0020] Step S202, using the first convolutional layer to identify local features of the time-frequency image, and determining feature dot products of each feature region to obtain a feature map of the time-frequency image;

[0021] Step S203, selecting a feature region with the largest feature point product in the 2×2 local region in the feature map as the 2×2 local region feature, and obtaining a local feature map;

[0022] Step S204, using the first fully connected layer to extract the local feature map into a one-dimensional feature vector, inputting the one-dimensional feature vector and the corresponding sample test capacitance into the output layer neurons for training, with the one-dimensional feature vector as input and the sample test capacitance as output, and defining the output sample test capacitance as sample ultrasonic capacitance, to obtain a first convolutional neural network model;

[0023] As a preferred embodiment of the lithium battery capacity detection method described in the present invention, wherein:

[0024] The step S3 specifically comprises the following steps:

[0025] Step S301, at a constant temperature T, a charge and discharge test is performed on the lithium battery to be tested, and an ultrasonic pulse signal is applied to the lithium battery using an ultrasonic transmitter to obtain ultrasonic feedback information to be tested;

[0026] Step S302, using the wavelet transform to obtain a time-frequency image to be measured according to the ultrasonic feedback information to be measured, and inputting the time-frequency image to be measured into a first convolutional neural network model to obtain the ultrasonic capacitance of the lithium battery to be measured;

[0027] As a preferred embodiment of the lithium battery capacity detection method described in the present invention, wherein:

[0028] The step S4 specifically comprises the following steps:

[0029] Step S401, obtaining the test capacity of the lithium battery to be tested according to the test capacity information, wherein the test capacity information includes a constant current and a discharge time;

[0030] Step S402, obtaining ultrasonic capacitance of ultrasonic capacitance information, wherein the ultrasonic capacitance information includes ultrasonic feedback wave information and ultrasonic capacitance of the lithium battery to be tested,

[0031] Step S403, performing weighted calculation on the test capacitance and the ultrasonic capacitance according to the test capacitance weight and the ultrasonic capacitance weight to obtain the capacitance of the lithium battery to be tested;

[0032] As a preferred embodiment of the lithium battery capacity detection method described in the present invention, wherein:

[0033] The step S5 specifically comprises the following steps:

[0034] Step S501, evaluating the health value of the lithium battery to be tested by using a battery health status formula according to the capacity of the lithium battery to be tested, and the calculation expression of the battery health status formula is as follows:

[0035] Among them, SOH is the health value of the lithium battery to be tested, C is the actual capacity of the lithium battery to be tested, is the rated capacity of the lithium battery to be tested, is the theoretical capacity of the lithium battery to be tested;

[0036] Step S502: if the health value is less than or equal to 80%, the lithium battery to be tested is defined as an unqualified lithium battery, and the internal ultrasonic image of the unqualified lithium battery is input into a battery fault model to obtain the battery fault of the unqualified lithium battery;

[0037] Establishing the battery failure model includes:

[0038] Acquire a battery fault data set, wherein the battery fault data set includes internal ultrasonic images of faulty batteries, data annotations, and fault labels of various faulty lithium batteries, input the internal ultrasonic images of the faulty batteries, the data annotations, and the fault labels into a second convolutional neural network model for training, with the faulty battery ultrasonic images as input and the fault labels as output, to acquire a battery fault model;

[0039] As a preferred embodiment of the lithium battery capacity detection method described in the present invention, wherein:

[0040] The use of wavelet transform to process feedback ultrasound includes:

[0041] Performing translation and scaling processing on the basic wavelet to obtain a wavelet basis function, and performing continuous wavelet transform on the feedback ultrasonic wave using the wavelet basis function to obtain a time-frequency image of the feedback ultrasonic wave;

[0042] The calculation expression for continuous wavelet transform is:

[0043] in, is the time-frequency of the feedback ultrasound after wavelet continuous transformation, is the scale factor, b is the translation factor, To feedback ultrasonic waves, is the conjugate function of the wavelet basis function; as a preferred embodiment of the lithium battery capacity detection method described in the present invention, wherein:

[0044] Obtaining the test capacitance weight and ultrasonic capacitance weight includes:

[0045] At a constant temperature T, the sample lithium battery is fully charged to reach the full-charge voltage, and then the sample lithium battery is discharged to reach the rated voltage, and the sample current and sample voltage of the sample lithium battery during the charging and discharging process are monitored in real time, and the sample current and sample voltage are integrated to obtain the sample charge and discharge capacity;

[0046] Obtaining a weight ratio according to the sample test capacitance, the sample charge-discharge capacitance and the sample ultrasonic capacitance by using a weight formula, wherein the weight ratio is a ratio of the test capacitance weight to the ultrasonic capacitance weight;

[0047] The calculation expression of the weight formula is as follows:

[0048] Among them, P is the weight ratio, n is the number of sample lithium batteries, is the sample test capacitance, is the sample charge and discharge capacity, is the ultrasonic capacitance of the sample.

[0049] In a second aspect, the present invention provides a lithium battery capacity detection system, including a data acquisition module, a capacity detection module and a fault diagnosis module;

[0050] The data acquisition module is used to obtain the test capacity information and ultrasonic capacity information of the lithium battery to be tested;

[0051] The capacity detection module is used to obtain the test capacity of the lithium battery to be tested according to the test capacity information, obtain the ultrasonic capacity of the lithium battery to be tested according to the ultrasonic capacity information, and obtain the capacity according to the test capacity and the ultrasonic capacity;

[0052] The fault diagnosis module is used to evaluate the health status of the lithium battery to be tested according to the capacitance, and diagnose the battery fault of the lithium battery to be tested through ultrasonic capacitance information according to the evaluation result.

[0053] In a third aspect, the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any one of the above-mentioned methods for detecting the capacity of a lithium battery when executing the computer program.

[0054] The beneficial effects of the present invention are as follows: the present invention combines ultrasonic technology with convolutional neural networks, utilizes wavelet transform to convert ultrasonic feedback information of a sample lithium battery into a time domain image, utilizes a convolutional neural network to obtain a first convolutional neural network model according to the time domain image, obtains the ultrasonic capacitance of the lithium battery to be tested through the first convolutional neural network model, utilizes a traditional charge and discharge test to obtain the test capacitance of the lithium battery to be tested, obtains the capacitance of the lithium battery to be tested based on the test capacitance and the ultrasonic capacitance, thereby improving the accuracy of lithium battery capacity detection and ensuring the accuracy of battery capacity detection.

[0055] Evaluating the health status of the lithium battery to be tested based on its battery capacity is beneficial to understanding the health status of the lithium battery to be tested in real time and ensuring the safe use of the lithium battery. In addition, obtaining the battery fault of the lithium battery to be tested using a fault model based on the internal ultrasonic image of the lithium battery to be tested is beneficial to timely handling of battery faults and ensuring the safe operation of the energy storage system. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A schematic diagram of a basic flow chart of a lithium battery capacity detection method provided by an embodiment of the present invention;

[0057] Figure 2 A schematic diagram of the basic flow of a lithium battery capacity detection system provided by one embodiment of the present invention; DETAILED DESCRIPTION

[0058] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0059] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a lithium battery capacity detection method, comprising the following steps:

[0060] Step S1, performing charge and discharge tests on sample lithium batteries respectively, obtaining sample test capacity information, and obtaining sample ultrasonic feedback information of each sample lithium battery respectively through ultrasonic detection technology, wherein T is a natural number;

[0061] Step S2, training a first convolutional neural network according to the sample ultrasound feedback information to obtain a first convolutional neural network model;

[0062] Step S3, performing a charge and discharge test on the lithium battery to be tested, obtaining test capacitance information and ultrasonic feedback information to be tested, and obtaining ultrasonic capacitance information by using the first convolutional neural network through the ultrasonic feedback information to be tested;

[0063] Step S4, obtaining the capacity of the lithium battery to be tested according to the test capacity information and the ultrasonic capacity information;

[0064] Step S5, evaluating the health status of the lithium battery to be tested according to the electric capacity, and obtaining the battery fault of the lithium battery by using a fault diagnosis model.

[0065] The charge and discharge test is an important test method for evaluating battery performance and capacity. It measures and analyzes the energy storage and release capabilities of the battery by simulating the charging and discharging process of the battery in actual use.

[0066] The convolutional neural network is a type of feedforward neural network that includes convolution calculations and has a deep structure, and is mainly used in image processing, target detection, and other computer vision fields.

[0067] In this embodiment, by combining ultrasonic technology and convolutional neural network, the ultrasonic feedback information of the sample lithium battery is converted into a time domain image by wavelet transform, and the convolutional neural network is used to obtain a first convolutional neural network model according to the time domain image, the ultrasonic capacitance of the lithium battery to be tested is obtained by the first convolutional neural network model, and the test capacitance of the lithium battery to be tested is obtained by traditional charge and discharge tests, and the capacitance of the lithium battery to be tested is obtained based on the test capacitance and the ultrasonic capacitance, thereby improving the accuracy of lithium battery capacity detection and ensuring the accuracy of battery capacity detection.

[0068] Evaluating the health status of the lithium battery to be tested based on its battery capacity is beneficial to understanding the health status of the lithium battery to be tested in real time and ensuring the safe use of the lithium battery. In addition, obtaining the battery fault of the lithium battery to be tested using a fault model based on the internal ultrasonic image of the lithium battery to be tested is beneficial to timely handling of battery faults and ensuring the safe operation of the energy storage system.

[0069] The step S1 specifically includes the following steps:

[0070] Step S101, randomly selecting n sample lithium batteries that have the same specifications as the lithium battery to be tested, where n is a natural number;

[0071] Step S102, at a constant temperature T, use a constant current to perform a charge and discharge test on n sample lithium batteries respectively, charge the sample lithium batteries with a constant current until the charge cut-off voltage is reached, then discharge the sample lithium batteries with a constant current until the discharge cut-off voltage is reached, and monitor the voltage of the sample lithium batteries in real time, and record the discharge time used to reach the cut-off voltage , the cut-off voltage includes a charge cut-off voltage and a discharge cut-off voltage;

[0072] Step S103, obtaining sample test capacity information of each sample lithium battery according to the constant current, discharge time and cut-off voltage, wherein the sample test capacity information includes the sample test capacity;

[0073] Step S104, using an ultrasonic transmitter to apply an ultrasonic pulse signal to the lithium battery, the ultrasonic pulse signal is sent to the ultrasonic receiving device after passing through the sample lithium battery, the ultrasonic receiving device transmits the acquired sample ultrasonic feedback information to the signal processing device, the sample ultrasonic feedback information includes charging ultrasonic feedback information and discharging ultrasonic feedback information, the sample ultrasonic feedback information specifically includes the internal ultrasonic image of the sample lithium battery and ultrasonic feedback wave information, the ultrasonic feedback wave information includes the waveform, amplitude and peak-to-peak value of the feedback ultrasonic wave.

[0074] In this embodiment, n is 100;

[0075] In this embodiment, the T is 25°C;

[0076] In this embodiment, the constant current is 0.2A.

[0077] The cut-off voltage is the voltage value at which the battery can no longer supply power or charge when the voltage drops to a certain level during the discharge and discharge process, indicating that the battery has been exhausted or fully charged;

[0078] The ultrasonic feedback wave is a sound wave with a frequency higher than 20KHZ;

[0079] The amplitude refers to the maximum peak value of the feedback ultrasonic wave in one detection cycle;

[0080] The peak-to-peak value refers to the difference between the maximum peak value and the minimum trough value of the feedback ultrasonic wave within a detection cycle;

[0081] In this embodiment, the sample capacity and sample ultrasonic feedback information of the sample lithium battery are obtained through charge and discharge tests, which provides detailed and specific training data for establishing the first convolutional neural network model and ensures the generalization of the model training.

[0082] The step S2 specifically includes the following steps:

[0083] Step S201, the signal processing device receives the sample ultrasonic feedback information, and processes the feedback ultrasonic wave by using a wavelet transform method to obtain a time-frequency image of the sample feedback ultrasonic wave;

[0084] Step S202, using the first convolutional layer to identify local features of the time-frequency image, and determining feature dot products of each feature region to obtain a feature map of the time-frequency image;

[0085] Step S203, selecting a feature region with the largest feature point product in the 2×2 local region in the feature map as the 2×2 local region feature, and obtaining a local feature map;

[0086] Step S204: Use the first fully connected layer to extract the local feature map into a one-dimensional feature vector, input the one-dimensional feature vector and the corresponding sample test capacitance into the output layer neurons for training, with the one-dimensional feature vector as input and the sample test capacitance as output, and define the output sample test capacitance as sample ultrasonic capacitance to obtain a first convolutional neural network model.

[0087] The time-frequency image refers to an image showing the changes in the frequency and amplitude of a signal over time.

[0088] In this embodiment, a wavelet transform method is used to obtain a time-frequency image, and a first convolutional neural network model is obtained according to the time-frequency image of the sample feedback ultrasound through a convolutional neural network, which provides an accurate model basis for obtaining the ultrasonic capacitance of the lithium battery to be tested.

[0089] The step S3 specifically comprises the following steps:

[0090] Step S301, at a constant temperature T, a charge and discharge test is performed on the lithium battery to be tested, and an ultrasonic pulse signal is applied to the lithium battery using an ultrasonic transmitter to obtain ultrasonic feedback information to be tested;

[0091] Step S302: using the wavelet transform to obtain a time-frequency image to be measured according to the ultrasonic feedback information to be measured, and inputting the time-frequency image to be measured into a first convolutional neural network model to obtain the ultrasonic capacitance of the lithium battery to be measured.

[0092] In this embodiment, the first convolutional neural network model is used to obtain the ultrasonic capacitance of the lithium battery to be tested according to the ultrasonic feedback information to be tested, which provides effective data support for obtaining the accurate capacitance of the lithium battery to be tested.

[0093] The step S4 specifically comprises the following steps:

[0094] Step S401, obtaining the test capacity of the lithium battery to be tested according to the test capacity information, wherein the test capacity information includes a constant current and a discharge time;

[0095] Step S402, obtaining ultrasonic capacitance of ultrasonic capacitance information, wherein the ultrasonic capacitance information includes ultrasonic feedback wave information and ultrasonic capacitance of the lithium battery to be tested,

[0096] Step S403, performing weighted calculation on the test capacitance and the ultrasonic capacitance according to the test capacitance weight and the ultrasonic capacitance weight to obtain the capacitance of the lithium battery to be tested.

[0097] In this embodiment, the capacity of the lithium battery to be tested is obtained according to the test capacity and the ultrasonic capacity, which improves the accuracy of the lithium battery capacity detection and ensures the accuracy of the battery capacity detection.

[0098] The step S5 specifically comprises the following steps:

[0099] Step S501, evaluating the health value of the lithium battery to be tested by using a battery health status formula according to the capacity of the lithium battery to be tested, and the calculation expression of the battery health status formula is as follows:

[0100] Among them, SOH is the health value of the lithium battery to be tested, C is the actual capacity of the lithium battery to be tested, is the rated capacity of the lithium battery to be tested, is the theoretical capacity of the lithium battery to be tested;

[0101] Step S502: if the health value is less than or equal to 80%, the lithium battery to be tested is defined as an unqualified lithium battery, and the internal ultrasonic image of the unqualified lithium battery is input into a battery fault model to obtain the battery fault of the unqualified lithium battery;

[0102] Establishing the battery failure model includes:

[0103] A battery failure data set is obtained, wherein the battery failure data set includes internal ultrasonic images of faulty batteries, data annotations, and fault labels of various faulty lithium batteries, and the internal ultrasonic images of the faulty batteries, the data annotations, and the fault labels are input into a second convolutional neural network model for training, with the faulty battery ultrasonic images as input and the fault labels as output, to obtain a battery failure model.

[0104] The battery failure includes gradual failure and sudden failure;

[0105] The theoretical capacity refers to the amount of electricity produced when all active substances participate in the battery reaction.

[0106] The actual capacity refers to the amount of electricity provided by the battery under certain discharge conditions;

[0107] The rated capacity refers to the capacity of the battery that can work continuously for a long time under rated working conditions;

[0108] In this embodiment, the health status of the lithium battery to be tested is evaluated according to the capacity of the lithium battery to be tested, and the battery fault of the unqualified lithium battery is obtained using the fault model, which is conducive to timely handling of battery faults and ensuring the safe operation of the energy storage system and the safety of battery use.

[0109] The use of wavelet transform to process feedback ultrasound includes:

[0110] Performing translation and scaling processing on the basic wavelet to obtain a wavelet basis function, and performing continuous wavelet transform on the feedback ultrasonic wave using the wavelet basis function to obtain a time-frequency image of the feedback ultrasonic wave;

[0111] The calculation expression for continuous wavelet transform is:

[0112] in, is the time-frequency of the feedback ultrasound after wavelet continuous transformation, is the scale factor, b is the translation factor, To feedback ultrasonic waves, is the conjugate function of the wavelet basis function.

[0113] The wavelet transform analyzes signals by scaling and translating basic wavelet functions, providing localized information in both time and frequency, and is suitable for processing non-stationary signals such as audio, images, and biomedical signals.

[0114] In this embodiment, the feedback ultrasound is converted into a time-frequency image using wavelet transform, which provides a clear source of time-frequency images for training the first convolutional neural network model and ensures the accuracy and generalization of the first convolutional neural network model.

[0115] Obtaining the test capacitance weight and ultrasonic capacitance weight includes:

[0116] At a constant temperature T, the sample lithium battery is fully charged to reach the full-charge voltage, and then the sample lithium battery is discharged to reach the rated voltage, and the sample current and sample voltage of the sample lithium battery during the charging and discharging process are monitored in real time, and the sample current and sample voltage are integrated to obtain the sample charge and discharge capacity;

[0117] Obtaining a weight ratio according to the sample test capacitance, the sample charge-discharge capacitance and the sample ultrasonic capacitance by using a weight formula, wherein the weight ratio is a ratio of the test capacitance weight to the ultrasonic capacitance weight;

[0118] The calculation expression of the weight formula is as follows:

[0119] Among them, P is the weight ratio, n is the number of sample lithium batteries, is the sample test capacitance, is the sample charge and discharge capacity, is the ultrasonic capacitance of the sample.

[0120] In this embodiment, a weight ratio is obtained based on the sample test capacitance, the sample charge and discharge capacitance, and the sample ultrasonic capacitance, which provides an accurate weight ratio for obtaining the capacitance of the lithium battery to be tested based on the test capacitance and the ultrasonic capacitance, thereby ensuring the accuracy of the capacitance.

[0121] Example 2, reference Figure 2 , which is another embodiment of the present invention, which is different from the first embodiment in that it provides a lithium battery capacity detection system, including a data acquisition module, a capacity detection module and a fault diagnosis module;

[0122] The data acquisition module is used to obtain the test capacity information and ultrasonic capacity information of the lithium battery to be tested;

[0123] The capacity detection module is used to obtain the test capacity of the lithium battery to be tested according to the test capacity information, obtain the ultrasonic capacity of the lithium battery to be tested according to the ultrasonic capacity information, and obtain the capacity according to the test capacity and the ultrasonic capacity;

[0124] The fault diagnosis module is used to evaluate the health status of the lithium battery to be tested according to the capacitance, and diagnose the battery fault of the lithium battery to be tested through ultrasonic capacitance information according to the evaluation result.

[0125] In this embodiment, by combining ultrasonic technology and convolutional neural network, the ultrasonic feedback information of the sample lithium battery is converted into a time domain image by wavelet transform, and the convolutional neural network is used to obtain a first convolutional neural network model according to the time domain image, the ultrasonic capacitance of the lithium battery to be tested is obtained by the first convolutional neural network model, and the test capacitance of the lithium battery to be tested is obtained by traditional charge and discharge tests, and the capacitance of the lithium battery to be tested is obtained based on the test capacitance and the ultrasonic capacitance, thereby improving the accuracy of lithium battery capacity detection and ensuring the accuracy of battery capacity detection.

[0126] Evaluating the health status of the lithium battery to be tested based on its battery capacity is beneficial to understanding the health status of the lithium battery to be tested in real time and ensuring the safe use of the lithium battery. In addition, obtaining the battery fault of the lithium battery to be tested using a fault model based on the internal ultrasonic image of the lithium battery to be tested is beneficial to timely handling of battery faults and ensuring the safe operation of the energy storage system.

[0127] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code.

[0128] -Wherein, the storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable programmable read only memory (EPROM), programmable red only memory (PROM), read only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including an instruction device, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0129] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A lithium battery capacity detection method, characterized in that: The following steps are involved: Step S1, performing charge and discharge tests on sample lithium batteries respectively, obtaining sample test capacity information, and obtaining sample ultrasonic feedback information of each sample lithium battery respectively through ultrasonic detection technology; Step S2, training a first convolutional neural network according to the sample ultrasound feedback information to obtain a first convolutional neural network model; Step S3, performing a charge and discharge test on the lithium battery to be tested, obtaining test capacitance information and ultrasonic feedback information to be tested, and obtaining ultrasonic capacitance information by using the first convolutional neural network through the ultrasonic feedback information to be tested; Step S4, obtaining the capacity of the lithium battery to be tested according to the test capacity information and the ultrasonic capacity information; Step S5, evaluating the health status of the lithium battery to be tested according to the electric capacity, and obtaining the battery fault of the lithium battery by using a fault diagnosis model; The step S2 specifically includes the following steps: Step S201, a signal processing device receives the sample ultrasonic feedback information, and processes the feedback ultrasonic wave using a wavelet transform method to obtain a time-frequency image of the sample feedback ultrasonic wave; Step S202, using the first convolutional layer to identify local features of the time-frequency image, and determining feature dot products of each feature region to obtain a feature map of the time-frequency image; Step S203, selecting a feature region with the largest feature point product in the 2×2 local region in the feature map as the 2×2 local region feature, and obtaining a local feature map; Step S204, using the first fully connected layer to extract the local feature map into a one-dimensional feature vector, inputting the one-dimensional feature vector and the corresponding sample test capacitance into the output layer neurons for training, with the one-dimensional feature vector as input and the sample test capacitance as output, and defining the output sample test capacitance as sample ultrasonic capacitance, to obtain a first convolutional neural network model; The step S3 specifically comprises the following steps: Step S301, at a constant temperature T, a charge and discharge test is performed on the lithium battery to be tested, and an ultrasonic pulse signal is applied to the lithium battery using an ultrasonic transmitter to obtain ultrasonic feedback information to be tested; Step S302: using the wavelet transform to obtain a time-frequency image to be measured according to the ultrasonic feedback information to be measured, and inputting the time-frequency image to be measured into a first convolutional neural network model to obtain the ultrasonic capacitance of the lithium battery to be measured.

2. A lithium battery capacity detection method as claimed in claim 1, characterized in that: The step S1 specifically includes the following steps: Step S101, randomly selecting n sample lithium batteries that have the same specifications as the lithium battery to be tested, where n is a natural number; Step S102, at a constant temperature T, use a constant current to perform a charge and discharge test on n sample lithium batteries respectively, charge the sample lithium batteries with a constant current until the charge cut-off voltage is reached, then discharge the sample lithium batteries with a constant current until the discharge cut-off voltage is reached, and monitor the voltage of the sample lithium batteries in real time, and record the discharge time used to reach the cut-off voltage , the cut-off voltage includes a charge cut-off voltage and a discharge cut-off voltage; Step S103, obtaining sample test capacity information of each sample lithium battery according to the constant current, discharge time and cut-off voltage, wherein the sample test capacity information includes the sample test capacity; Step S104, using an ultrasonic transmitter to apply an ultrasonic pulse signal to the lithium battery, the ultrasonic pulse signal is sent to the ultrasonic receiving device after passing through the sample lithium battery, the ultrasonic receiving device transmits the acquired sample ultrasonic feedback information to the signal processing device, the sample ultrasonic feedback information includes charging ultrasonic feedback information and discharging ultrasonic feedback information, the sample ultrasonic feedback information specifically includes the internal ultrasonic image of the sample lithium battery and ultrasonic feedback wave information, the ultrasonic feedback wave information includes the waveform, amplitude and peak-to-peak value of the feedback ultrasonic wave.

3. A lithium battery capacity detection method as claimed in claim 1, characterized in that: The step S4 specifically comprises the following steps: Step S401, obtaining the test capacity of the lithium battery to be tested according to the test capacity information, wherein the test capacity information includes a constant current and a discharge time; Step S402, obtaining ultrasonic capacitance of ultrasonic capacitance information, wherein the ultrasonic capacitance information includes ultrasonic feedback wave information and ultrasonic capacitance of the lithium battery to be tested, Step S403, performing weighted calculation on the test capacitance and the ultrasonic capacitance according to the test capacitance weight and the ultrasonic capacitance weight to obtain the capacitance of the lithium battery to be tested.

4. A lithium battery capacity detection method as claimed in claim 1, characterized in that: The step S5 specifically comprises the following steps: Step S501, evaluating the health value of the lithium battery to be tested by using a battery health status formula according to the capacity of the lithium battery to be tested, and the calculation expression of the battery health status formula is as follows: Among them, SOH is the health value of the lithium battery to be tested, C is the actual capacity of the lithium battery to be tested, is the rated capacity of the lithium battery to be tested, is the theoretical capacity of the lithium battery to be tested; Step S502: if the health value is less than or equal to 80%, the lithium battery to be tested is defined as an unqualified lithium battery, and an internal ultrasonic image of the unqualified lithium battery is input into a battery fault model to obtain a battery fault of the unqualified lithium battery; Establishing the battery failure model includes: A battery failure data set is obtained, wherein the battery failure data set includes internal ultrasonic images of faulty batteries, data annotations, and fault labels of various faulty lithium batteries, and the internal ultrasonic images of the faulty batteries, the data annotations, and the fault labels are input into a second convolutional neural network model for training, with the internal ultrasonic images of the faulty batteries as input and the fault labels as output, to obtain a battery failure model.

5. A lithium battery capacity detection method as claimed in claim 3, characterized in that: The use of wavelet transform to process feedback ultrasound includes: Performing translation and scaling processing on the basic wavelet to obtain a wavelet basis function, and performing continuous wavelet transform on the feedback ultrasonic wave using the wavelet basis function to obtain a time-frequency image of the feedback ultrasonic wave; The calculation expression for continuous wavelet transform is: in, is the time-frequency of the feedback ultrasound after wavelet continuous transformation, is the scale factor, b is the translation factor, To feedback ultrasonic waves, is the conjugate function of the wavelet basis function.

6. A lithium battery capacity detection method as claimed in claim 4, characterized in that: Obtaining the test capacitance weight and ultrasonic capacitance weight includes: At a constant temperature T, the sample lithium battery is fully charged to reach the full-charge voltage, and then the sample lithium battery is discharged to reach the rated voltage, and the sample current and sample voltage of the sample lithium battery during the charging and discharging process are monitored in real time, and the sample current and sample voltage are integrated to obtain the sample charge and discharge capacity; Obtaining a weight ratio according to the sample test capacitance, the sample charge-discharge capacitance and the sample ultrasonic capacitance by using a weight formula, wherein the weight ratio is a ratio of the test capacitance weight to the ultrasonic capacitance weight; The calculation expression of the weight formula is as follows: Among them, P is the weight ratio, n is the number of sample lithium batteries, is the sample test capacitance, is the sample charge and discharge capacity, is the ultrasonic capacitance of the sample.

7. A lithium battery capacity detection system, characterized in that: It includes a data acquisition module, a capacitance detection module and a fault diagnosis module; The data acquisition module is used to obtain the test capacity information and ultrasonic capacity information of the lithium battery to be tested; The capacity detection module is used to obtain the test capacity of the lithium battery to be tested according to the test capacity information, obtain the ultrasonic capacity of the lithium battery to be tested according to the ultrasonic capacity information, and obtain the capacity according to the test capacity and the ultrasonic capacity; The fault diagnosis module is used to evaluate the health status of the lithium battery to be tested according to the capacitance, and diagnose the battery fault of the lithium battery to be tested through ultrasonic capacitance information according to the evaluation result; Among them, the ultrasonic capacitance information is obtained in the following way: The signal processing device receives the sample ultrasonic feedback information, and processes the feedback ultrasonic wave by using the wavelet transform method to obtain the time-frequency image of the sample feedback ultrasonic wave; Using the first convolutional layer to identify local features of the time-frequency image, and determining feature point products of each feature region, to obtain a feature map of the time-frequency image; Selecting a feature region with the largest feature point product in the 2×2 local region in the feature map as the 2×2 local region feature, and obtaining a local feature map; The local feature map is extracted into a one-dimensional feature vector by using the first fully connected layer, and the one-dimensional feature vector and the corresponding sample test capacitance are input into the output layer neurons for training, with the one-dimensional feature vector as input and the sample test capacitance as output, and the output sample test capacitance is defined as the sample ultrasonic capacitance, to obtain a first convolutional neural network model; At a constant temperature T, a charge and discharge test is performed on the lithium battery to be tested, and an ultrasonic transmitter is used to apply an ultrasonic pulse signal to the lithium battery to obtain ultrasonic feedback information to be tested; The wavelet transform is used to obtain a time-frequency image to be measured according to the ultrasonic feedback information to be measured, and the time-frequency image to be measured is input into a first convolutional neural network model to obtain the ultrasonic capacitance of the lithium battery to be measured.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for detecting the capacity of a lithium battery according to any one of claims 1 to 6 is implemented.

Images