Battery service life testing method and system

Abstract

The invention relates to the technical field of battery testing, and particularly discloses a battery service life testing method and system, and the method comprises the steps: obtaining testing condition information; generating test condition information corresponding to the test condition information by using the CGAN model; converting the test condition information into test control information; and testing the service life of the to-be-tested battery according to the test control information. The test working condition information corresponding to the test condition information is generated by using the CGAN model, the test working condition information produced by the CGAN model approaches to the real working condition, and the service life of the battery to be tested is tested by using the test control information converted from the test working condition information. Therefore, the test environment and the test current in the life test process tend to the real driving environment and the real driving charging and discharging current. The CGAN model can simulate the complex and changeable use conditions of the new energy vehicle, so that the to-be-tested battery simulates the actual driving charge-discharge process of the new energy vehicle in the test process, thereby ensuring that the service life of the to-be-tested battery can be tested more accurately.

Description

Technical Field

[0001] The present invention relates to the technical field of battery testing, and particularly to a method and system for testing the service life of a battery. Background Art

[0002] As one of the important alternatives to traditional fuel vehicles, new energy vehicles have gradually entered people's lives. Among all the components of new energy vehicles, the battery system is one of the most important and core components, directly determining the service life of new energy vehicles. By conducting a constant current and constant voltage cycling life test on the battery, it helps to evaluate the performance, life, and safety of the battery to ensure that the battery can provide reliable power output in applications. However, due to the complex and variable working conditions of new energy vehicles in actual use, it is difficult to accurately evaluate the service life of the battery in the application scenarios of new energy vehicles using traditional constant current and constant voltage charging and constant current discharging cycling life tests. Summary of the Invention

[0003] Based on this, in view of the problem that it is difficult to accurately evaluate the service life of the battery in the application scenarios of new energy vehicles, it is necessary to provide a method and system for testing the service life of the battery.

[0004] A method for testing the service life of a battery includes obtaining test condition information; generating test working condition information corresponding to the test condition information by using a CGAN model; converting the test working condition information into test control information; and performing a life test on the battery to be tested according to the test control information.

[0005] In one of the embodiments, the converting the test working condition information into test control information includes: Calculating the speed working condition in the test working condition information as a battery current working condition according to a conversion formula; Wherein, the conversion formula includes: In the formula, represents the motor power, in kW; represents the vehicle transmission efficiency; represents the vehicle rolling resistance coefficient; represents the vehicle's own gravity, in N; represents the vehicle speed, in km / h; represents the road slope; represents the vehicle's aerodynamic drag coefficient; represents the vehicle's frontal area, in m 2 ; represents the air density, 1.2258 kg / m 3 ; represents the instantaneous acceleration, in m / s 2 ; represents the vehicle weight, in kg; represents the rotating mass exchange coefficient; In the formula, represents the battery system discharge current, in A; represents the motor voltage, in V.

[0006] In one embodiment, the converting the test condition information into test control information further includes: converting the battery current condition into a condition file in a preset format.

[0007] In one embodiment, the performing a life test on the battery under test according to the test control information includes: controlling the test current and test environment applied to the battery under test during the test according to the condition file.

[0008] In one embodiment, the generating the test condition information corresponding to the test condition information by using the CGAN model further includes: constructing and training the CGAN model, wherein the constructing and training the CGAN model includes obtaining the historical real condition information of different types of vehicles; preprocessing the historical real condition information to form a training data set of sample data; constructing the CGAN model based on the CGAN neural network algorithm, and the CGAN model includes a model input unit, a conditional generator model unit, a real sample input unit, and a discriminator model unit; training the CGAN model based on the training data set until the CGAN model converges.

[0009] In one embodiment, the preprocessing the historical real condition information to form a training data set of sample data includes normalizing the historical real condition information by using the Max-Min normalization calculation method to form the training data set; wherein, the Max-Min normalization calculation method is: In the formula, represents the sample data at time t in the training data set, represents the data at time t in the historical real condition information before normalization processing, represents the minimum value in the historical real condition information, represents the maximum value in the historical real condition information.

[0010] In one embodiment, training the CGAN model based on the training data set until the CGAN model converges includes initializing the conditional generator model unit and the discriminator model unit; repeating the training process according to the pattern of training the discriminator model unit every preset number of times and then training the conditional generator model unit once until the CGAN model converges; wherein, the convergence of the CGAN model includes the parameters of the discriminator model unit being stable, and the parameters of the conditional generator model unit being stable.

[0011] In one embodiment, training the conditional generator model unit includes: Inputting the sample data and random noise into the conditional generator model unit; The conditional generator model unit extracts features from the conditional information of the sample data and the random noise, and generates test condition information according to Equation (1); and Updating the parameters of the conditional generator model unit according to Equation (2) ; wherein, Equation (1) is: In Equation (1), z represents random noise, c represents conditional information, represents the test condition information generated according to the random noise under the conditional information, represents a recurrent network implemented using Relu as the activation function; wherein, Equation (2) is: In Equation (2), represents the relevant parameters of the conditional generator model unit, represents the probability that the test condition information is real data, S represents that the sample is real, represents the test condition information, and N represents the number of samples.

[0012] In one embodiment, training the discriminator model unit includes: Inputting the sample data and the test condition information into the discriminator model unit; The discriminator model unit compares the sample data and the test condition information, and outputs the probability that the test condition information is real data according to Equation (3); and Updating the parameters of the discriminator model unit according to Equation (4) ; wherein, Equation (3) is: In formula (3), represents the probability that the test condition information is real data, represents the function implemented by the network hidden layer of the discriminator model unit, represents the activation function, S represents that the sample is real, represents the test condition information; wherein, formula (4) is: In the formula, represents the relevant parameters of the discriminator model unit, represents the real condition sample data, represents the unreal condition sample data, S represents that the sample is real, represents that the sample is not real, represents the test condition information, represents the non-test condition information, and N represents the number of samples.

[0013] A battery service life test system is used to execute the battery service life test method of any one of the above embodiments. The battery service life test system includes a control module for obtaining test condition information and converting the test condition information into test control information. Wherein, the test condition information is generated based on the test condition information by using a CGAN model; a test module connected to the control module for adjusting the test environment of the battery under test according to the test control information and performing a life test on the battery under test; a monitoring module connected to the test module for obtaining test data of the battery under test during the life test and determining the battery attenuation condition of the battery under test according to the test data.

[0014] In one embodiment, the control module includes a working condition generation unit configured to obtain the test condition information and generate test working condition information corresponding to the test condition information by using a pre-trained CGAN model; a working condition confirmation unit connected to the working condition generation unit and configured to determine whether the test working condition information meets the requirements, and in response to a judgment result that the test working condition information meets the requirements, convert the test working condition information into test control information; a working condition control unit connected to the working condition confirmation unit and configured to control the charge and discharge cabinet to charge and discharge the battery under test according to the test control information, and control the climate chamber to adjust the test environment of the battery under test according to the test control information; the test module includes a charge and discharge cabinet connected to the battery under test and the working condition control unit and configured to perform charge and discharge tests on the battery under test; a climate chamber connected to the working condition control unit and configured to provide different test environments for the battery under test; a sensor group connected to the battery under test and configured to monitor battery monitoring data of the battery under test during the charge and discharge process; the monitoring module includes a data acquisition unit connected to the sensor group and configured to acquire the battery monitoring data; a data monitoring unit connected to the data acquisition unit and configured to determine the battery attenuation condition of the battery under test according to the battery monitoring data.

[0015] In one embodiment, the monitoring module further includes a display screen connected to the data monitoring unit and configured to display the battery monitoring data; an audible and visual alarm device connected to the data monitoring unit and configured to alarm or give early warning of abnormal conditions of the battery under test.

[0016] The above battery service life test method uses a CGAN model to generate test working condition information corresponding to the test condition information. The test working condition information generated by the CGAN model is close to the real working condition. The test control information converted from the test working condition information is used to test the service life of the battery under test, so that the test environment and test current during the life test process also tend to be the real driving environment and real driving charge and discharge current. During the test, the working state of the battery under test is continuously monitored to determine the attenuation condition of the battery under test. The CGAN model can simulate the complex and changeable use conditions of new energy vehicles, so that the battery under test simulates the actual driving charge and discharge process of new energy vehicles during the test, thereby ensuring that the service life test of the battery under test can be carried out more accurately. Description of the Drawings

[0017] To more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments recorded in this specification. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a schematic flowchart of the battery service life test method in one embodiment of this application; Figure 2 It is a schematic flowchart of the method for constructing and training the CGAN model in one embodiment of this application; Figure 3 It is a schematic structural diagram of the CGAN model in one embodiment of this application; Figure 4 It is a schematic flowchart of the method for training the CGAN model in one embodiment of this application; Figure 5 It is a schematic structural diagram of the battery service life test system in one embodiment of this application; Figure 6 It is a schematic structural diagram of the battery service life test system in another embodiment of this application; Figure 7 It is a schematic diagram of the speed condition curve output by the CGAN model in one embodiment of this application; Figure 8 It is a schematic diagram of the battery capacity attenuation curve in one embodiment of this application; Figure 9 It is a schematic structural diagram of the system for implementing the battery service life test method in one embodiment of this application; Figure 10 It is an internal structural diagram of a computer device in one embodiment of this application. Detailed implementation manners

[0019] To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant accompanying drawings. The preferred embodiments of the present invention are given in the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to understand the disclosure of the present invention more thoroughly and comprehensively.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs. The terms used in the specification of this invention are for the purpose of describing specific embodiments only and are not intended to limit the invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0021] It should also be noted that the information (including but not limited to test condition information, historical actual working condition information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0022] This application provides a method for testing the service life of a battery based on a CGAN neural network, which can automatically generate working condition data of a new energy vehicle's battery under specified usage conditions, thereby simulating the usage of a lithium-ion battery under actual working conditions and improving the accuracy of testing the service life of a lithium battery. In this application, CGAN is an abbreviation for Conditional Generative Adversarial Network (conditional generative neural network).

[0023] In some embodiments, the method for testing the service life of a battery can be executed by a battery service life test system. For example, the method for testing the service life of a battery can be stored in the test system in the form of a program or instruction, and when the program or instruction is executed, the method for testing the service life of a battery can be implemented. The device disclosed in this application for implementing the method for testing the service life of a battery can be either a device with a large amount of computing resources (such as a computer, a server, cloud computing, etc.) or a device with limited computing resources (such as a hardware circuit such as an FPGA chip board or an ASIC chip board).

[0024] Figure 1 This is a schematic flowchart of the method for testing the service life of a battery in one embodiment of this application. In one embodiment, the method for testing the service life of a battery can include the following steps S100 to step S400.

[0025] Step S100: Obtain test condition information.

[0026] In some embodiments, before testing the service life of the battery to be tested, the test condition information of the battery to be tested in actual application can be input through the interactive operation between the operator and the system. Or in some embodiments, different driving scenarios can be preset in the system, and the test condition information in this service life test can be determined by the operator's selection or random setting.

[0027] Based on the test condition information, the usage situation of the battery under test in actual application can be simulated. For example, according to the test condition information, it is determined that the battery under test is assembled on a new energy vehicle for an actual driving scenario. Lithium batteries are often used in new energy vehicles due to their high energy density, long cycle life, and low self-discharge rate, etc., to provide power for them. Therefore, in this embodiment, the battery under test can refer to a lithium battery, which is used to store and provide the electric power required for the operation of the new energy vehicle.

[0028] In some embodiments, the test condition information is common usage information of new energy vehicles, such as it can include the driving city, driving distance, driving section, and so on. The tester can directly reasonably select the test condition information according to the scenario of the battery under test being put into use in the vehicle, or obtain relevant test condition information from the market side.

[0029] In some specific embodiments, the test condition information can include but is not limited to information such as single-trip driving mileage, driving condition, driving location, number of traffic lights, etc. For example, the single-trip driving mileage is set to 20 km, the driving condition is set to urban condition, the driving location is set to Suzhou, the number of traffic lights is set to medium number of traffic lights, and the vehicle type is set to medium-sized sedan. That is, this service life test simulates the battery operation condition of a medium-sized new energy sedan driving 20 km in Suzhou city with a medium number of traffic lights.

[0030] Step S200: Use the CGAN model to generate test condition information corresponding test working condition information.

[0031] Take the test condition information as the input information of the CGAN model, and use the CGAN model to output the corresponding test working condition information according to the test condition information. Among them, the CGAN model can be a prediction model pre-constructed and trained based on the CGAN neural network algorithm, and the CGAN model can be used to generate test working condition information under specified usage conditions (i.e., under the test condition information). , where M is the number of generated test working conditions.

[0032] In this embodiment, the test working condition information can refer to the running state of the vehicle under specified usage conditions, including but not limited to the vehicle speed, ambient temperature, driving time, driving distance, etc. For example, when this service life test simulates a small new energy sedan driving 30 km in the case of a large number of traffic lights, the CGAN model can generate test working condition information simulating the change of the vehicle speed with time during the whole running process according to the above test condition information.

[0033] In some embodiments, the CGAN model can generate M test working conditions based on the test condition information. , M is an integer greater than 1; in some specific embodiments, to meet the test requirements, the CGAN model can generate multiple test conditions based on the test condition information, and the tester selects the test conditions that meet the test needs from the multiple test conditions.

[0034] Step S300: Convert the test condition information into test control information.

[0035] Convert the test condition information into test control information for the charge and discharge cabinet and the climate chamber to read. In this embodiment, the test control information may refer to the control information used to control the operation of the charge and discharge cabinet and the climate chamber.

[0036] Step S400: Perform a life test on the battery under test according to the test control information.

[0037] By converting the test condition information into test control information, test equipment such as the charge and discharge cabinet and the climate chamber can provide corresponding charge and discharge processes and corresponding test environments to the battery under test under the control of the test control information, and carry out a service life test on the battery under test.

[0038] The battery service life test method based on the CGAN neural network provided by this application uses a CGAN model pre-constructed and trained based on the CGAN neural network algorithm to generate test condition information corresponding to test condition information. The test condition information produced by the CGAN model approaches the real working conditions. The test control information converted from the test condition information is used to test the service life of the battery under test, so that the test environment and test current during the life test also tend to the real driving environment and real driving charge and discharge current. During the test, continuously monitor the working state of the battery under test to determine the attenuation of the battery under test. The CGAN model can simulate the complex and changeable usage conditions of new energy vehicles, so that the battery under test simulates the actual driving charge and discharge process of new energy vehicles during the test, thereby ensuring that the service life test of the battery under test can be carried out more accurately.

[0039] In one embodiment, after using the CGAN model to generate the test condition information corresponding to the test condition information, it is also possible to determine whether the test condition information meets the requirements.

[0040] The CGAN model can also determine whether the test condition information it generates meets the requirements, that is, determine whether the generated test condition information is real. Specifically, the CGAN model generates multiple test condition information under specified usage conditions, and can determine whether the test condition information meets the requirements based on the judgment result of whether the generated test condition information is real, and select the test condition information with the judgment result of real to carry out the service life test on the battery under test.

[0041] When the test condition information meets the requirements, it can be determined to use the test condition information generated this time for testing, and the test condition information is converted into test control information for easy reading by the charge and discharge cabinet and the climate chamber. In this embodiment, the test control information may refer to the control information for controlling the operation of the charge and discharge cabinet and the climate chamber.

[0042] In one embodiment, when the test condition information does not meet the requirements, the CGAN model is controlled to regenerate the test condition information under specified conditions, and the CGAN model is trained again until the test condition information output by the CGAN model meets the requirements.

[0043] In one embodiment, according to the test control information, the life test of the battery under test may specifically include the following steps: controlling the charge and discharge process of the battery under test according to the test control information, and adjusting the test environment of the battery under test according to the test control information.

[0044] By converting the test condition information into test control information, the charge and discharge cabinet and the climate chamber can provide the corresponding charge and discharge process and the corresponding test environment to the battery under test under the control of the test control information.

[0045] In one embodiment, in the life test of the battery under test, the following steps may further be included: monitoring the battery under test to obtain the battery monitoring data of the battery under test; determining the battery attenuation condition of the battery under test according to the battery monitoring data.

[0046] After obtaining the test control information, the life test of the battery under test can be carried out. During the test, the battery under test is continuously monitored to obtain the battery monitoring data of the battery under test. The battery attenuation condition of the battery under test is determined according to the battery monitoring data. At the same time, the state of the battery under test can also be monitored according to the battery monitoring data to ensure that the battery under test is in a safe state.

[0047] In one embodiment, step S300 may specifically include calculating the speed condition in the test condition information as the battery current condition according to the conversion formula. Further, the test control information can be compiled into a file in a specified form, and the battery current condition is converted into a working condition file in a preset format for the control programs of the charge and discharge cabinet and the climate chamber to read.

[0048] Among them, the conversion formula may include: In the formula, represents the motor power, unit kW; represents the vehicle transmission efficiency; represents the vehicle rolling resistance coefficient; represents the vehicle's own gravity, unit N; represents the vehicle speed, unit km / h; represents the road gradient; represents the vehicle aerodynamic drag coefficient; represents the vehicle frontal area, unit m 2 ; represents the air density, 1.2258 kg / m 3 ; represents the instantaneous acceleration, unit m / s 2 ; represents the vehicle weight, unit kg; represents the rotating mass exchange coefficient; wherein, the road gradient is the ratio h / l of the height difference h (unit m) between the starting point and the ending point of the vehicle travel to the horizontal distance l (unit m) between the starting point and the ending point.

[0049] In the formula, represents the current released by the battery system, unit A; represents the motor voltage, unit V.

[0050] In one of the embodiments, according to the test control information, the life test of the battery under test may further include the following steps: controlling the test current and test environment applied to the battery under test during the test according to the working condition file. Test equipment such as charge and discharge cabinets and climate environmental test chambers can provide the corresponding charge and discharge process and corresponding test environment to the battery under test by reading the working condition file.

[0051] Figure 2 is a schematic flow chart of the method for constructing and training the CGAN model in one of the embodiments of the present application. In one of the embodiments, using the CGAN model to generate the test working condition information corresponding to the test condition information further includes: constructing and training the CGAN model, wherein the method for constructing and training the CGAN model may include the following steps S210 to step S240.

[0052] Step S210: Obtain the historical real working condition information of different types of vehicles.

[0053] Collect the historical real working condition information of different types of vehicles to construct a data set according to the historical real working condition information. The historical real working condition information may include but is not limited to information such as user portraits, weather conditions, current data, driving mileage, etc., wherein the user portrait may include but is not limited to information such as gender, age, job position, driving preference, regional characteristics, etc.

[0054] Step S220: Preprocess the historical real working condition information to form a training data set of sample data.

[0055] Preprocess the obtained historical real operating condition information to form the training dataset X of sample data. The preprocessing of historical real operating condition information mainly includes censoring of information and processing of abnormal data, and data normalization can also be performed.

[0056] Step S230: Construct a CGAN model based on the CGAN neural network algorithm. The CGAN model includes a model input unit, a conditional generator model unit, a real sample input unit, and a discriminator model unit.

[0057] Figure 3 It is a schematic structural diagram of the CGAN model in one embodiment of the present application. As Figure 3 shown, the CGAN model mainly includes a model input unit, a conditional generator model unit G, a real sample input unit, and a discriminator model unit D. Among them, the real sample input unit can use the method of random sampling to extract data samples from the training dataset X of sample data, and input the extracted data samples into the discriminator model unit D. The model input unit can input the conditional information c corresponding to the data samples extracted by the real sample input unit and the random noise z generated by the model input unit into the conditional generator model unit G. The conditional generator model unit G can generate test operating condition information I(z|c) according to the input conditional information c and random noise z. The discriminator model unit D can determine the probability that the test operating condition information I(z|c) generated by the conditional generator model unit G is real data.

[0058] Step S240: Train the CGAN model based on the training dataset until the CGAN model converges.

[0059] Use the training dataset X of sample data to train the CGAN model. The conditional generator model unit G generates test operating condition information according to the given conditions, and the discriminator model unit D determines whether the generated test operating condition information is real. The conditional generator model unit G adjusts its own model parameters according to the determination result of the discriminator, and the discriminator model unit D is trained according to the real sample data and the generated test operating condition information data to optimize the parameters of the discriminator model unit D. The conditional generator model unit G and the discriminator model unit D are trained according to the set training mode until the CGAN model converges.

[0060] By training the CGAN model with the generated data and the real data together, the performance of the CGAN model can be optimized, and the generalization ability and prediction accuracy of the model can be improved. In the service life test of the battery to be measured, the CGAN model is used to automatically generate the test condition information corresponding to the test working conditions. Conducting the test based on the test working conditions generated by the CGAN model can more accurately simulate the operation of the battery to be measured in the actual application scenario of new energy vehicles, and thus can accurately evaluate the service life of the battery in the application scenario of new energy vehicles.

[0061] In one embodiment, preprocessing the historical real working condition information to form a training data set of sample data may include the following steps: using the Max-Min normalization calculation method to normalize the historical real working condition information to form a training data set X.

[0062] Among them, the Max-Min normalization calculation method may be: In the formula, represents the sample data x at time t in the training data set X, represents the data at time t in the historical real working condition information before normalization, represents the minimum value in the historical real working condition information, represents the maximum value in the historical real working condition information. In a specific implementation manner, xmin may be 0 km / h, and xmax may be (120 km / h * 120%) = 144 km / h.

[0063] In one embodiment, the conditional generator model unit G may include a network hidden layer and a network output layer. Among them, the network hidden layer of the conditional generator model unit G may be used to extract features from the conditional information c and the random noise z, and the extracted features are determined by the conditional generator model unit G itself; the network output layer of the conditional generator model unit G may be used to generate the test working condition information I(z|c) according to the conditional information c and the random noise z after feature extraction.

[0064] Among them, the calculation method of the battery working condition may be: In formula (1), z represents the random noise, c represents the conditional information, represents the test working condition information generated according to the random noise z under the conditional information c, represents a recurrent network implemented using Relu as the activation function.

[0065] In one embodiment, the discriminator model unit D may include an input layer, a network hidden layer, and a network output layer. Among them, the input layer of the discriminator model unit D may be used to take the data sample x and the test condition information I(z|c) generated by the conditional generator model unit G as inputs, and the output layer is used to output the probability P(S|I) that the test condition information I(z|c) is real data according to the probability calculation method.

[0066] Among them, the probability calculation method may be: In Equation (3), represents the probability that the test condition information is real data, represents the function implemented by the network hidden layer of the discriminator model unit, represents the activation function, S represents that the sample is real, represents the test condition information.

[0067] Figure 4 FIG. 24 is a schematic flowchart of a method for training a CGAN model in one embodiment of the present application. In one embodiment, training the CGAN model based on the training data set until the CGAN model converges may include the following steps S241 to step S243.

[0068] Step S241: Initialize the conditional generator model unit and the discriminator model unit.

[0069] Step S243: Repeat the training process according to the mode of training the conditional generator model unit once every preset number of times of training the discriminator model unit until the CGAN model converges.

[0070] After initializing the parameters of the conditional generator model unit G and the discriminator model unit D, repeat the training process according to the mode of training the conditional generator model unit G once every preset number of times of training the discriminator model unit D until the CGAN model converges. Among them, the preset number of times may also be set when initializing the parameters. For example, it may be 5, that is, after training the discriminator model unit D 5 times, train the conditional generator model unit G once.

[0071] Among them, the convergence of the CGAN model may include the parameters of the discriminator model unit being stable, and the parameters of the conditional generator model unit being stable.

[0072] In one embodiment, training the conditional generator model unit may include the following steps: Input the sample data and random noise into the conditional generator model unit.

[0073] The conditional generator model unit extracts features from the conditional information and random noise of the sample data, and generates test condition information according to Equation (1). And update the parameters of the conditional generator model unit according to Equation (2) .

[0074] Among them, Equation (1) is:[[]]END]] In Equation (1), z represents random noise, c represents conditional information,[[]]END]] represents the test condition information generated according to the random noise z under the conditional information c,[[]]END]] represents a recurrent network implemented using Relu as the activation function.[[]]END]]

[0075] Among them, Equation (2) is:[[]]END]] In Equation (2),[[]]END]] represents the relevant parameters of the conditional generator model unit G,[[]]END]] represents the probability that the test condition information is real data, S represents the sample as real,[[]]END]] represents the test condition information, and N represents the number of samples.[[]]END]]

[0076] Specifically, in this field, the parameter[[]]END]] is the model parameter of the conditional generator model[[]]END]] is the set of weights and biases in this model. Similar to the parameter update of other network models, the parameter[[]]END]] is randomly generated by the conditional generator model according to a predefined strategy or a given strategy.[[]]END]]

[0077] Specifically, when training the conditional generator model unit G, the network hidden layer of the conditional generator model unit G extracts features from the conditional information c and random noise z, and the network output layer of the conditional generator model unit G generates test condition information I(z|c) according to the conditional information c and random noise z after feature extraction. The input layer of the discriminator model unit D takes the data sample x and the test condition information I(z|c) generated by the conditional generator model unit G as inputs, and the output layer of the discriminator model unit D outputs the probability P(S|I) that the test condition information I(z|c) is real data according to the probability calculation method, that is, the discriminator model unit D determines whether the data generated by the conditional generator model unit G is real.[[]]END]]

[0078] After each training of the conditional generator model unit G is completed, the conditional generator model unit G can adjust the relevant parameters of the conditional generator model unit G according to the feedback P(S|I) of the discriminator model unit D.[[]]END]]

[0079] In one of the embodiments, training the discriminator model unit may include the following steps:[[]]END]] Input the sample data and test condition information into the discriminator model unit; The discriminator model unit compares the sample data and the test condition information, and outputs the probability that the test condition information is real data according to Equation (3); and Update the parameters of the discriminator model unit according to Equation (4). ; Among them, Equation (3) is: In Equation (3), represents the test condition information is the probability of real data, represents the function implemented by the network hidden layer of the discriminator model unit D, represents the activation function, S represents that the sample is real, represents the test condition information; Among them, in the formula, represents the relevant parameters of the discriminator model unit D, which is a set of weights and biases in the discriminator model, and is also randomly generated by the discriminator model according to the established strategy or the given strategy.

[0080] Equation (4) is: In the formula, represents the relevant parameters of the discriminator model unit D, represents the real condition sample data, represents the unreal condition sample data, S represents that the sample is real, represents that the sample is not real, represents the test condition information, represents the non-test condition information, and N represents the number of samples.

[0081] Specifically, represents the probability of determining the generated test condition information as real, represents the probability of determining the real condition sample data as real; represents the probability of determining the generated unreal test condition information as real; represents the probability of determining the real condition sample data as not real. By making as large as possible through optimization, avoid the situation that the discriminator model unit D cannot accurately distinguish the difference between the model-generated condition and the real condition data, that is, avoid the situation of being unable to distinguish the model-generated condition and the real condition.

[0082] When training the discriminator model unit D, the discriminator model unit D outputs the probability P(S|I) that the test condition information I(z|c) is real data according to the real data of the training data set X and the test condition information I(z|c) generated by the conditional generator model unit G, that is, to determine whether the data generated by the conditional generator model unit G is real. After each training of the discriminator model unit D is completed, the discriminator model unit D can update the relevant parameters of the discriminator model unit D according to the real data and the generated data 。

[0083] It should be understood that although the steps in the flowchart of the accompanying drawings of the specification are shown in sequence according to the indication of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise clearly stated in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings of the specification may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or steps or stages in other steps

[0084] Based on the description of the above embodiments of the battery service life test method, the present disclosure also provides a battery service life test system. The system may include a system (including a distributed system), software (application), module, component, server, client, etc. that uses the method described in the embodiments of the present specification and combines the necessary implementation hardware. Based on the same innovative concept, the system in one or more embodiments provided by the embodiments of the present disclosure is as described in the following embodiments. Since the implementation solutions for the system to solve problems are similar to those of the method, the implementation of the specific system in the embodiments of the present specification can refer to the implementation of the foregoing method, and the repeated parts will not be described again. As used hereinafter, the term "unit" or "module" may be a combination of software and / or hardware that can implement a predetermined function. Although the systems described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated

[0085] The present application also provides a battery service life test system, Figure 5 which is a schematic structural diagram of the battery service life test system in one embodiment of the present application. In one embodiment, the battery service life test system may include a control module 100, a test module 200, and a monitoring module 300

[0086] The control module 100 can be used to generate test control information according to test condition information. The system can also include a multimedia component for providing an interaction means between the system and the operator. For example, devices such as a touch screen, a mouse, a keyboard, buttons, etc., to receive information or instructions input by the operator. The operator can input the test condition information of the battery under test in actual application through the interaction operation with the system. Or, different form scenarios can be preset in the system, and the test condition information in this service life test is determined by the operator's selection or random setting method.

[0087] After obtaining the test condition information, the control module 100 can use the test condition information as the input information of the CGAN model, and use the CGAN model to output the corresponding test working condition information according to the test condition information. A prediction model constructed and trained based on the CGAN neural network algorithm can be configured in the control module 100. The control module 100 can also use the CGAN model to judge whether the generated test working condition information meets the requirements, that is, judge whether the generated test working condition information is real. When the test working condition information meets the requirements, it can be determined to use the test working condition information generated this time for testing. The control module 100 converts the test working condition information into test control information for the test module 200 to read.

[0088] The test module 200 can be connected to the battery under test 10 and the control module 100. The test module 200 can be used to control the charge and discharge process of the battery under test 10 according to the test control information, and adjust the test environment of the battery under test 10 during the charge and discharge process according to the test control information. The test module 200 can provide the corresponding charge and discharge process and the corresponding test environment to the battery under test 10 under the control of the test control information and perform a life test on the battery under test 10.

[0089] The monitoring module 300 can be connected to the test module 200. The monitoring module 300 can be used to monitor the battery under test 10 during the test, obtain the test data of the battery under test 10 during the life test, and determine the battery attenuation situation of the battery under test 10 according to the test data. During the service life test, the monitoring module 300 continuously monitors the battery under test 10 to obtain the test data of the battery under test 10. The monitoring module 300 can determine the battery attenuation situation of the battery under test 10 according to the test data. At the same time, it can also monitor the state of the battery under test 10 according to the test data to ensure that the battery under test is in a safe state.

[0090] Using the battery service life test system provided by this application to test the battery to be tested, the control module 100 can use the CGAN model to generate test working condition information corresponding to the test condition information, and the control module 100 uses the test control information converted from the test working condition information to control the test module 200 to implement the service life test for the battery 10 to be tested. During the test, the monitoring module 300 continuously monitors the working state of the battery 10 to be tested to determine the attenuation condition of the battery 10 to be tested. The CGAN model can simulate the complex and changeable working conditions of new energy vehicles, so as to ensure that the system can more accurately carry out the service life test on the battery to be tested.

[0091] Figure 6 FIG. 4 is a schematic structural diagram of the battery service life test system in another embodiment of this application. In one embodiment, the control module 100 may include a working condition generation unit 110, a working condition confirmation unit 120, and a working condition control unit 130. The test module 200 may include a charge and discharge cabinet 210, a climate environment test chamber 220, and a sensor group 230. The monitoring module 300 may include a data acquisition unit 310 and a data monitoring unit 320.

[0092] The working condition generation unit 110 may be used to obtain test condition information and generate test working condition information corresponding to the test condition information by using the CGAN model. Specifically, the working condition generation unit 110 may use the conditional generator model unit G of the CGAN model to generate test working condition information corresponding to the test condition information , where M is the number of generated working conditions. The test working condition information may include information such as the driving speed of the vehicle.

[0093] The working condition confirmation unit 120 may be connected to the working condition generation unit 110. The working condition confirmation unit 120 may be used to determine whether the test working condition information meets the requirements. Specifically, the working condition confirmation unit 120 may use the discriminator model unit D of the CGAN model to determine the probability that the test working condition information output by the conditional generator model unit G is real data.

[0094] In one embodiment, when the working condition confirmation unit 120 determines that the test working condition information does not meet the requirements, it may control the CGAN model to regenerate the test working condition information under specified conditions and retrain the CGAN model until the test working condition information output by the CGAN model meets the requirements.

[0095] The operating condition confirmation unit 120 can also be used to convert the test operating condition information into test control information in response to the judgment result that the test operating condition information meets the requirements. Specifically, the operating condition confirmation unit 120 can calculate the speed operating condition in the test operating condition information as the battery current operating condition according to the conversion formula. Further, the operating condition confirmation unit 120 can also compile the test control information into a file in a specified format for the charge and discharge cabinet and the control program of the climate chamber to read.

[0096] Among them, the conversion formula may include: In the formula, represents the motor power, unit kW; represents the vehicle transmission efficiency; represents the vehicle rolling resistance coefficient; represents the vehicle's own gravity, unit N; represents the vehicle speed, unit km / h; represents the road gradient; represents the vehicle aerodynamic drag coefficient; represents the vehicle frontal area, unit m 2 ; represents the air density, 1.2258 kg / m 3 ; represents the instantaneous acceleration, unit m / s 2 ; represents the vehicle weight, unit kg; represents the rotating mass exchange coefficient; In the formula, represents the battery system discharge current, unit A; represents the motor voltage, unit V.

[0097] The operating condition control unit 130 can be connected to the operating condition confirmation unit 120. The operating condition control unit 130 controls the charge and discharge cabinet 210 and the climate chamber 220 to test the battery under test 10 according to the test control information. The operating condition control unit 130 can be used to adjust the battery charge and discharge current data of the charge and discharge cabinet 210 under specified usage conditions according to the test control information to implement the charge and discharge process for the battery under test. The operating condition control unit 130 controls the climate chamber 220 to adjust the test environment of the battery under test 10 according to the test control information.

[0098] The charge and discharge cabinet 210 can be connected to the battery under test 10 and the operating condition control unit 130. The charge and discharge cabinet 210 can be used to perform charge and discharge tests on the battery under test 10. The charge and discharge cabinet 210 applies a specified current to the battery under test 10 according to the test control information.

[0099] The climate environment test chamber 220 can be connected to the working condition control unit 130. The climate environment test chamber 220 can be used to provide different test environments for the battery under test 10, that is, the climate environment test chamber 220 can adjust the external environments such as the ambient temperature, ambient humidity, and air pressure of the battery under test 10 to simulate the real application environment of the battery under test 10. The climate environment test chamber 220 applies a specified test environment to the battery under test 10 according to the test control information.

[0100] The sensor group 230 can be connected to the battery under test 10. The sensor group 230 can be used to monitor the battery monitoring data of the battery under test 10 during the charge and discharge process. In a specific implementation manner, a series of sensors can be installed inside or outside the battery under test 10 to monitor the state of the battery under test 10 during the test and ensure that the battery under test 10 is in a safe state. The installed sensors include but are not limited to an internal temperature sensor, an external temperature sensor, a pressure sensor, a current sensor, a voltage sensor, and an air pressure sensor.

[0101] The data acquisition unit 310 can be connected to the sensor group 230. The data acquisition unit 310 can be used to acquire the battery monitoring data. The data acquisition module 310 acquires the data monitored by the sensor group 230 installed on the battery under test 10.

[0102] The data monitoring unit 320 can be connected to the data acquisition unit 310. The data monitoring module 320 can be responsible for processing the data acquired by the data acquisition module 310 and determining the battery attenuation condition of the battery under test 10 according to the battery monitoring data.

[0103] In one embodiment, the monitoring module 300 can further include a display screen 330 and an audible and visual alarm device 340.

[0104] The display screen 330 can be connected to the data monitoring unit 320. The display screen 330 can be used to display the battery monitoring data. After the data monitoring unit 320 completes the processing of the data acquired by the data acquisition module 310, the data monitoring unit 320 can display the acquired data on the display screen 330.

[0105] The acoustic-optic alarm device 340 can be connected to the data monitoring unit 320. The data monitoring module 320 can also automatically determine whether the current battery under test 10 is in a normal state or predict whether the battery under test 10 is about to have an abnormality by analyzing data changes and other means, and control the acoustic-optic alarm device 340 to alarm or give an early warning about the abnormal situation of the battery under test 10. For example, when the battery under test 10 is in an abnormal state, the data monitoring module 320 controls the acoustic-optic alarm device 340 to emit red light for alarm; when the battery under test 10 is about to have an abnormality, the data monitoring module 320 controls the acoustic-optic alarm device 340 to emit orange light for alarm.

[0106] In this embodiment, information such as "urban driving condition, driving mileage of 20 Km, medium-sized sedan, medium number of traffic lights, Suzhou" is used as the test condition information to describe the above battery service life test system based on the CGAN neural network. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent.

[0107] The driving condition generation unit 110 generates test driving condition information corresponding to the input test condition information by using the conditional generator model unit G of the CGAN model obtained through pre-training. The driving condition confirmation unit 120 uses the discriminator model unit D of the CGAN model to judge the probability that the test driving condition information output by the conditional generator model unit G is real data. When the driving condition confirmation unit 120 determines that the test driving condition information meets the requirements, the service life test of the battery under test 10 can be carried out based on the test driving condition information. In this embodiment, Figure 7 is a schematic diagram of the speed driving condition curve output by the CGAN model in one embodiment of the present application. As Figure 7 shown, in the test driving condition information output by the CGAN model, the speed driving condition curve changes with the simulated vehicle adopting different driving speeds at different times under the test conditions.

[0108] The driving condition confirmation unit 120 can convert the speed driving condition in the test driving condition information into a battery current driving condition according to the conversion formula, that is, convert the simulated speed-time curve into a current-time curve. For example, at time t, the test driving condition information specifically includes a vehicle rolling resistance coefficient of 0.02; the vehicle's own gravity of 19600 N; the road gradient of 0; the vehicle transmission efficiency of 0.92; the vehicle aerodynamic drag coefficient of 0.34; the vehicle frontal area of 1.5 m 2 ; the vehicle weight of 1960 kg; the rotating mass exchange coefficient of 1.12; when the vehicle speed When the speed is 40 km / h, the acceleration is 1 m / s 2 and other information. Based on the above test condition information, substituting into the motor power calculation gives: At the same time, the vehicle battery system is 2P80S, so the current is From this, the current data required for the life test is calculated, and the control program of the charge and discharge cabinet 210 and the climate chamber 220 is generated using the working condition confirmation module 120 to read the file. The charge and discharge cabinet 210 and the climate chamber 220 can apply the specified current and external environment to the battery under test 10 according to this file. The data acquisition module 310 acquires the data monitored by a series of sensors installed on the battery under test 10, the data monitoring module 320 processes the data acquired by the data acquisition module 310, and displays the acquired data on the display screen 330. The data monitoring module 320 automatically determines whether the current battery under test 10 is normal and predicts whether the battery under test 10 will have an abnormality. In the case of an abnormal condition, the data monitoring module 320 controls the sound and light alarm device 340 to output the corresponding sound and light alarm. The system repeats the above process until the input test conditions are met.

[0109] At the same time, during the test process, the system can also perform a cell capacity test every 7 days. Figure 8 This is a schematic diagram of the battery capacity attenuation curve in one embodiment of the present application. Figure 8 It shows the battery capacity attenuation curves drawn by using the above battery life test system to perform life tests on test cell 1 and test cell 2 respectively, with a cell capacity test every 7 days and based on the results of the cell capacity test.

[0110] It can be understood that each embodiment of the above methods, systems, etc. in this specification is described in a progressive manner. The same / similar parts between each embodiment can be referred to each other, and the key points of each embodiment are the differences from other embodiments. For the relevant parts, refer to the description of other method embodiments.

[0111] Figure 9 This is a schematic diagram of the system structure for implementing the battery life test method in one embodiment of the present application. Refer to Figure 9, the battery life test system S00 may include a processing component S20, which further includes one or more processors, and memory resources represented by a memory S22 for storing instructions executable by the processing component S20, such as application programs. The application programs stored in the memory S22 may include one or more instructions, each module corresponding to a set of instructions. In addition, the processing component S20 is configured to execute instructions to perform the above battery life test method.

[0112] The battery life test system S00 may further include: a power supply component S24 configured to perform power management of the battery life test system S00, a wired or wireless network interface S26 configured to connect the battery life test system S00 based on the CGAN neural network to a network, and an input / output (I / O) interface S28. The battery life test system S00 may operate based on an operating system stored in the memory S22, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, or the like.

[0113] In an exemplary embodiment, there is also provided a computer-readable storage medium including instructions, such as the memory S22 including instructions, and the above instructions can be executed by the processor of the battery life test system S00 to complete the above method. The storage medium may be a computer-readable storage medium. For example, the computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0114] In an exemplary embodiment, there is also provided a computer program product, and the computer program product includes instructions, and the above instructions can be executed by the processor of the battery life test system S00 to complete the above method.

[0115] In one embodiment, there is provided a computer device, and the computer device may be a server, and its internal structure diagram may be as Figure 10 shown Figure 10The following is the internal structure diagram of a computer device in one embodiment of the present application. The computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data related to users and tasks used in the above battery service life test method. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer program is executed by the processor to implement a battery service life test method.

[0116] Those skilled in the art can understand that Figure 10 the structure shown in is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0117] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0118] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. The key point of each embodiment is to illustrate the differences from other embodiments. In particular, for the hardware + program type embodiments, since they are basically similar to the method embodiments, they are described relatively simply. For the relevant parts, reference can be made to the partial description of the method embodiments.

[0119] It should be noted that the above-described devices, electronic devices, servers, etc. may also include other implementation manners according to the description of the method embodiments. The specific implementation manners may refer to the description of the relevant method embodiments. At the same time, new embodiments formed by the mutual combination of the features among the various methods, as well as the device, equipment, and server embodiments, still fall within the scope of the embodiments covered by the present disclosure, and will not be elaborated one by one here.

[0120] In the description of this specification, the description with reference to terms such as "some embodiments", "other embodiments", "ideal embodiments", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic description of the above terms does not necessarily refer to the same embodiment or example.

[0121] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as falling within the scope described in this specification.

[0122] The above-described embodiments only represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the invention patent should be subject to the appended claims.

Claims

1. A battery life testing method, characterized in that: include: Get test condition information; Generate test condition information corresponding to the test condition information using the CGAN model; Converting the test condition information into test control information; A life test is performed on the battery to be tested according to the test control information.

2. The battery life testing method based on CGAN neural network according to claim 1, characterized in that: The converting the test condition information into test control information comprises: Calculate the speed condition in the test condition information into the battery current condition according to a conversion formula; Wherein, the conversion formula includes: In the formula, Indicates motor power, unit kW; Indicates vehicle transmission efficiency; Indicates the rolling resistance coefficient of the vehicle; Indicates the vehicle's own gravity, in N; Indicates vehicle speed in km / h; Indicates the road slope; Indicates the drag coefficient of the car; Indicates the frontal area of ​​the vehicle, in m 2 ; Indicates air density, 1.2258kg / m 3 ; Indicates instantaneous acceleration, unit: m / s 2 ; Indicates the vehicle weight in kg; represents the rotational mass exchange coefficient; In the formula, Indicates the current released by the battery system, in A; Indicates the motor voltage, in V.

3. The battery life testing method according to claim 1 or 2, characterized in that: The converting the test operating condition information into test control information also includes: converting the battery current operating condition into an operating condition file in a preset format.

4. The battery life testing method according to claim 3, characterized in that: The performing a life test on the battery to be tested according to the test control information includes: controlling a test current and a test environment applied to the battery to be tested during the test according to the operating condition file.

5. The battery life testing method according to claim 1, characterized in that: The step of generating the test condition information corresponding to the test condition information by using the CGAN model further includes: constructing and training the CGAN model, wherein the step of constructing and training the CGAN model includes: Obtain historical real working condition information of different types of vehicles; Preprocessing the historical real working condition information to form a training data set of sample data; Constructing the CGAN model based on the CGAN neural network algorithm, wherein the CGAN model includes a model input unit, a condition generator model unit, a real sample input unit and a discriminator model unit; The CGAN model is trained based on the training data set until the CGAN model converges.

6. The battery life testing method according to claim 5, characterized in that: The preprocessing of the historical real working condition information to form a training data set of sample data includes: The historical real working condition information is normalized by using the Max-Min normalization calculation method to form the training data set; The Max-Min normalization calculation method is: In the formula, represents the sample data in the training data set at time t, represents the data at time t in the historical real working condition information before normalization, represents the minimum value in the historical real working condition information, Indicates the maximum value of the historical real operating condition information.

7. The battery life testing method according to claim 5, characterized in that: The step of training the CGAN model based on the training data set until the CGAN model converges includes: Initializing the conditional generator model unit and the discriminator model unit; Repeat the training process in a mode of training the conditional generator model unit once for each preset number of training of the discriminator model unit until the CGAN model converges; The CGAN model converges to the parameters of the discriminator model unit. stable, and the parameters of the conditional generator model unit Stablize.

8. The battery life testing method according to claim 7, characterized in that: Training the conditional generator model unit includes: Inputting the sample data and random noise into the condition generator model unit; The condition generator model unit extracts features from the condition information of the sample data and the random noise, and generates test condition information according to formula (1); and Update the parameters of the conditional generator model unit according to formula (2) ; Among them, formula (1) is: In formula (1), z represents random noise, c represents conditional information, represents the test condition information generated according to the random noise under the condition information, Represents a recursive network implemented using Relu as the activation function; Among them, formula (2) is: In formula (2), represents the relevant parameters of the condition generator model unit, represents the probability that the test condition information is real data, S represents that the sample is real, represents the test condition information, and N represents the number of samples.

9. The battery life testing method according to claim 8, characterized in that: Training the discriminator model unit includes: Inputting the sample data and the test condition information into the discriminator model unit; The discriminator model unit compares the sample data with the test condition information, and outputs the probability that the test condition information is true data according to formula (3); and Update the parameters of the discriminator model unit according to formula (4) ; Wherein, the formula (3) is: In formula (3), Indicates the probability that the test condition information is real data, represents the function implemented by the network hidden layer of the discriminator model unit, represents the activation function, S represents the sample is real, Indicates the test condition information; Wherein, the formula (4) is: In the formula, represents the relevant parameters of the discriminator model unit, represents the real working condition sample data, represents the sample data of unreal working condition, S represents the sample is real, Indicates that the sample is not real. Indicates the test condition information, represents information of non-described test conditions, and N represents the number of samples.

10. A battery life test system, used to execute the battery life test method according to any one of claims 1 to 9, characterized in that: The battery life test system comprises: A control module, used to obtain test condition information and convert the test condition information into test control information, wherein the test condition information is generated based on the test condition information using a CGAN model; A test module, connected to the control module, for adjusting the test environment of the battery to be tested according to the test control information and performing a life test on the battery to be tested; The monitoring module is connected to the testing module and is used to obtain the test data of the battery to be tested during the life test process, and determine the battery attenuation of the battery to be tested according to the test data.

Images