A rapid testing system and method for the cycle life of lithium-ion batteries

Abstract

This invention proposes a rapid cycle life testing system and method for lithium-ion batteries, relating to the field of functional testing, and more specifically to the field of electrical performance testing. The system includes: multiple analysis mechanisms for acquiring various radio electromagnetic test data, correlation parameters, and the average cycle life and previous batch intervals of lithium-ion batteries from each historical production batch; and a cycle life assessment mechanism for intelligently assessing the cycle life of the current lithium-ion battery based on the analysis results from the multiple analysis mechanisms using an intelligent cycle life assessment model. This invention addresses the technical problem of the difficulty in rapidly and accurately testing the cycle life of massive quantities of lithium-ion batteries one by one. By introducing a customized intelligent cycle life assessment model and comprehensively selecting multiple fundamental information sources, an intelligent cycle life numerical analysis mechanism is constructed, thereby solving the aforementioned technical problem.

Description

Technical Field

[0001] The functional testing of this invention relates more specifically to the field of electrical performance testing, and more specifically to a rapid testing system and method for the cycle life of lithium-ion batteries. Background Technology

[0002] Lithium-ion batteries are rechargeable batteries that primarily function by the movement of lithium ions between the positive and negative electrodes. During charging and discharging, lithium ions repeatedly insert and extract between the two electrodes: during charging, lithium ions extract from the positive electrode, pass through the electrolyte, and insert into the negative electrode, leaving the negative electrode in a lithium-rich state; the reverse occurs during discharging. The cycle life of lithium-ion batteries varies significantly due to differences in material systems and technologies. Mainstream products typically have a cycle life exceeding 1600-4000 cycles. Key influencing factors include material chemical properties, charge / discharge management strategies, and ambient temperature control. While lithium-ion batteries of the same model and batch may have a similar range of cycle life values, the specific values ​​can still vary considerably between individual batteries of the same model and batch.

[0003] For example, Chinese invention patent publication CN108732499A proposes a method and system for detecting the cycle life of lithium-ion batteries. This method includes detecting the cycle life of lithium-ion batteries under the same temperature and atmospheric pressure conditions using different current rates, and plotting the relationship curve between the number of cycles and the state parameters of the lithium-ion battery. Based on a constructed equivalent model of battery cycle life under different current rates, the cycle life of lithium-ion batteries under standard testing at low current rates is rapidly deduced using high current rates. The technical solution provided by this invention effectively shortens the cycle life detection time of lithium-ion batteries and improves the accuracy of lithium-ion battery life prediction.

[0004] For example, Chinese invention patent publication CN107748338A discloses a device and method for detecting and evaluating the cycle life of lithium-ion batteries. This device includes two base plates and a pressure sensor. The two base plates are arranged in parallel, with the lithium-ion battery under test located between them. The pressure sensor is mounted on the side of one base plate opposite to the lithium-ion battery, and its output is electrically connected to a signal processor. Because the pressure sensor is located on the base plate, during charge-discharge cycle testing, the pressure sensor promptly reflects the pressure changes caused by the expansion of the lithium-ion battery in the thickness direction to the signal processor. By analyzing the relationship between the number of cycles, capacity decay rate, and pressure changes, the cycle life of the lithium-ion battery can be evaluated.

[0005] Therefore, the existing technical solutions mentioned above are only limited to the extrapolation or evaluation of the cycle life of lithium-ion batteries. The resulting cycle life values ​​of lithium-ion batteries differ significantly from the actual cycle life values, and the accuracy is not high. At the same time, the cycle life value analysis mechanism, which is too biased towards physical measurement and numerical extrapolation, requires a lot of computing, time and storage costs, and the analysis rate is slow, which cannot meet the needs of rapid one-by-one testing of the cycle life of a large number of lithium-ion batteries. Summary of the Invention

[0006] To address the technical problems in existing technologies, this invention provides a rapid cycle life detection system and method for lithium-ion batteries. Based on a customized intelligent cycle life assessment model and a comprehensive selection of fundamental information, it replaces the overly physical measurement and numerical extrapolation-based cycle life numerical analysis mechanism with an artificial intelligence model-based mechanism. This improves the detection accuracy and speed of lithium-ion battery cycle life while reducing the computational, time, and storage costs of cycle life numerical analysis, thus meeting the need for rapid, one-by-one testing of the cycle life of large batches of lithium-ion batteries.

[0007] According to one aspect of the present invention, a rapid cycle life testing system for lithium-ion batteries is provided, the system comprising:

[0008] The first analysis unit is used to obtain the internal resistance data, capacity data, and open circuit voltage data of the current lithium-ion battery after it has been tested by radio electromagnetic testing when it is in the factory state, so as to use as multiple radio electromagnetic test data of the current lithium-ion battery.

[0009] The second analysis unit is used to obtain the current lithium-ion battery's volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity as various related parameters of the current lithium-ion battery.

[0010] The third analysis unit is used to take the current lithium-ion battery production batch as the target production batch, and obtain the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch and the interval between each previous batch. The number of lithium-ion batteries produced in each production batch is equal.

[0011] An object assembly mechanism is used to perform multiple trainings on a feedforward neural network to obtain a feedforward neural network after multiple trainings and output it as a cycle lifetime intelligent identification model.

[0012] The lifespan assessment mechanism is connected to the first analysis mechanism, the second analysis mechanism, the third analysis mechanism, and the object assembly mechanism, respectively. It is used to intelligently assess the cycle life of the current lithium-ion battery using a cycle life intelligent assessment model based on multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch, as well as the interval between each previous batch.

[0013] According to another aspect of the present invention, a method for rapid detection of the cycle life of a lithium-ion battery is provided, the method comprising:

[0014] The internal resistance data, capacity data, and open-circuit voltage data obtained from the radio electromagnetic testing of the current lithium-ion battery at the factory state are used as multiple radio electromagnetic test data of the current lithium-ion battery.

[0015] Obtain the current lithium-ion battery's volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity as various related parameters of the current lithium-ion battery;

[0016] The current lithium-ion battery production batch is taken as the target production batch. The average cycle life of each lithium-ion battery and the interval between each previous batch are obtained for each historical production batch of the same model before the target production batch. The number of lithium-ion batteries produced in each production batch is equal.

[0017] The feedforward neural network is trained multiple times to obtain a feedforward neural network after multiple trainings, which is then used as the output of the cycle lifetime intelligent identification model.

[0018] The cycle life intelligent assessment model uses multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, the average cycle life of each lithium-ion battery in each historical production batch of the same model before the target production batch, and the interval between each previous batch to intelligently assess the cycle life of the current lithium-ion battery.

[0019] Therefore, it can be seen that the present invention has at least the following prominent substantive features:

[0020] Substantial Feature A: Intelligent identification of the average cycle life of each lithium-ion battery in its factory condition is performed based on an artificial intelligence model. The intelligent identification is based on radio electromagnetic test data, lithium-ion battery correlation parameters, and the average cycle life of lithium-ion batteries of the same model in various historical batches, thereby completing the rapid detection of the cycle life of each lithium-ion battery in its factory condition, improving the speed and accuracy of lithium-ion battery cycle life detection.

[0021] Substantive Feature B: To intelligently determine the cycle life of a current lithium-ion battery in its factory condition, a customized intelligent cycle life assessment model is introduced. This model is a feedforward neural network that has been trained multiple times. The number of training iterations of the feedforward neural network follows the numerical trend of the number of lithium-ion batteries produced in each production batch. The feedforward neural network consists of an output layer, an input layer, and multiple hidden layers. The multiple hidden layers are located between the output layer and the input layer, and the number of hidden layers follows the numerical trend of the batch numbers of previous historical production batches of the same model before the target production batch. The customized structural design of the above-mentioned intelligent cycle life assessment model ensures the stability and effectiveness of the intelligent assessment results of the cycle life of the current lithium-ion battery in its factory condition.

[0022] Substantial Feature C: To intelligently determine the cycle life of a current lithium-ion battery in its factory condition, multiple pieces of basic information are introduced. These pieces of basic information include multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery in each historical production batch of the same model before the target production batch, as well as the interval between each previous batch. The comprehensive selection of the above-mentioned basic information further ensures the stability and effectiveness of the intelligent determination result of the cycle life of the current lithium-ion battery in its factory condition.

[0023] Substantial Feature D: The current lithium-ion battery's multiple radio electromagnetic test data include internal resistance data, capacity data, and open-circuit voltage data obtained after radio electromagnetic testing. The current lithium-ion battery's various related parameters include its volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity. Furthermore, taking the current lithium-ion battery's production batch as the target production batch, the average cycle life of each lithium-ion battery from each historical production batch of the same model prior to the target production batch and the interval between each previous batch are obtained. The number of lithium-ion batteries produced in each production batch is equal, thus completing the customized design of the data structure for the input information used for intelligent identification.

[0024] Substantive Feature E: In each training iteration of the feedforward neural network, the known cycle life of a lithium-ion battery that has reached the end of its lifespan is used as a single output of the feedforward neural network. Meanwhile, multiple radio electromagnetic test data of the lithium-ion battery, various correlation parameters of the lithium-ion battery, and the average cycle life of lithium-ion batteries from each historical production batch preceding the lithium-ion battery's production batch are used as sequential inputs to the feedforward neural network to complete the training. This ensures the effectiveness of each training iteration of the feedforward neural network. Attached Figure Description

[0025] The embodiments of the present invention will now be described with reference to the accompanying drawings, wherein:

[0026] Figure 1 This is a schematic diagram of the working scenario of a rapid cycle life testing system and method for lithium-ion batteries according to the present invention.

[0027] Figure 2 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a first embodiment of the present invention.

[0028] Figure 3 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a second embodiment of the present invention.

[0029] Figure 4 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a third embodiment of the present invention.

[0030] Figure 5 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a fourth embodiment of the present invention.

[0031] Figure 6 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a fifth embodiment of the present invention.

[0032] Figure 7 The present invention provides a flowchart of a method for rapid testing of the cycle life of a lithium-ion battery according to a sixth embodiment of the present invention. Detailed Implementation

[0033] like Figure 1 The diagram illustrates a working scenario of a rapid cycle life testing system and method for lithium-ion batteries according to the present invention. The functional testing of this invention specifically relates to the field of electrical performance testing.

[0034] exist Figure 1 To achieve this invention, the following technical processes are specifically proposed:

[0035] Technical Process 1: To intelligently determine the cycle life of a lithium-ion battery in its factory condition, a customized intelligent cycle life assessment model is introduced, such as... Figure 1 As shown;

[0036] Specifically, the customized structural design of the intelligent cycle life assessment model is mainly reflected in the following aspects:

[0037] First: The cycle life intelligent identification model is a feedforward neural network that has been trained multiple times. The feedforward neural network consists of an output layer, an input layer and multiple hidden layers, with the multiple hidden layers located between the output layer and the input layer.

[0038] Second: The number of training iterations of the feedforward neural network used follows the numerical trend of the number of lithium-ion batteries produced in each production batch.

[0039] For example, when each production batch produces an equal number of 10,000 lithium-ion batteries, the selected feedforward neural network undergoes 200 training iterations; when each production batch produces an equal number of 20,000 lithium-ion batteries, the selected feedforward neural network undergoes 400 training iterations; when each production batch produces an equal number of 30,000 lithium-ion batteries, the selected feedforward neural network undergoes 600 training iterations, and so on.

[0040] Therefore, through the numerical analysis of the above training times, intelligent cycle life identification models with different structures can be constructed for production batches of different sizes.

[0041] Third: In the feedforward neural network used, the number of hidden layers follows the trend of batch number changes of each historical production batch of the same model before the target production batch.

[0042] Fourth: In each training iteration of the feedforward neural network, the known cycle life of a lithium-ion battery that has reached the end of its lifespan is used as the single output of the feedforward neural network. The multiple radio electromagnetic test data of the lithium-ion battery, the various correlation parameters of the lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch before the production batch of the lithium-ion battery are used as the input of the feedforward neural network to complete the training. This ensures the training effect of the feedforward neural network in each training iteration.

[0043] In this way, through the customized structural design of the cycle life intelligent assessment model, the stability and effectiveness of the intelligent assessment results of the current cycle life of lithium-ion batteries under factory conditions are ensured.

[0044] Technical Process 2: To intelligently determine the cycle life of the current lithium-ion battery in its factory condition, several basic information items are introduced;

[0045] Specifically, the aforementioned basic information includes multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries of the same model prior to the target production batch, such as... Figure 1 As shown, the average cycle life of each lithium-ion battery from each historical production batch of the same model preceding the target production batch and the interval between each preceding batch are respectively: Figure 1 Two historical batch information items;

[0046] More specifically, the current radio electromagnetic test data for lithium-ion batteries includes internal resistance data, capacity data, and open-circuit voltage data obtained by radio electromagnetic testing of current lithium-ion batteries after standing.

[0047] More specifically, the relevant parameters of current lithium-ion batteries are the current lithium-ion battery's volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity;

[0048] More specifically, the current lithium-ion battery production batch is taken as the target production batch, and the average cycle life of each lithium-ion battery and the interval between each previous batch are obtained for each historical production batch of the same model before the target production batch. The number of lithium-ion batteries produced in each production batch is equal.

[0049] In this way, based on the customized design of the data structure for the input information used for intelligent identification, the full and comprehensive selection of multiple basic information is realized, which further ensures the stability and effectiveness of the intelligent identification results of the current cycle life of lithium-ion batteries in the factory condition.

[0050] Technical Process 3: Using the intelligent cycle life assessment models with different structures built for different production batches of different sizes in Technical Process 1, and based on the multiple basic information fully and comprehensively selected in Technical Process 2, the intelligent assessment of the cycle life of the current lithium-ion battery in its factory condition is completed.

[0051] In this way, a cycle life numerical analysis mechanism based on an artificial intelligence model was adopted to replace the cycle life numerical analysis mechanism that was too biased towards physical measurement and numerical extrapolation.

[0052] Technical Process 4: If the cycle life of the current lithium-ion battery obtained from the intelligent identification in Technical Process 3 is lower than the minimum cycle life required to meet the factory requirements, the current lithium-ion battery will be marked as a defective battery to prevent it from entering the market.

[0053] Conversely, if the cycle life of the current lithium-ion battery obtained by intelligent identification in its factory condition is higher than or equal to the minimum cycle life required to meet the factory requirements, the current lithium-ion battery will be marked as a qualified battery and allowed to enter the market.

[0054] Therefore, through the coordinated operation of the above-mentioned technical processes, the specific cycle life numerical analysis of each lithium-ion battery based on the artificial intelligence model has been realized. This replaces the cycle life numerical analysis mechanism that is too biased towards physical measurement and numerical extrapolation, improves the detection accuracy and detection rate of lithium-ion battery cycle life, and reduces the computational cost, time cost and storage cost of cycle life numerical analysis, thus meeting the need for rapid one-by-one detection of the cycle life of a large number of lithium-ion batteries.

[0055] The key points of this invention are: the effective replacement of the cycle life numerical analysis mechanism based on artificial intelligence model with the cycle life numerical analysis mechanism that is too biased towards physical measurement and numerical inference; the customized construction of intelligent cycle life identification models for different structures of different production batches of different scales; the full and comprehensive selection of multiple basic information for intelligent numerical analysis of cycle life; and the targeted design of each training of the feedforward neural network.

[0056] The present invention will now be described in detail by way of embodiments of a rapid detection system and method for the cycle life of lithium-ion batteries.

[0057] First Embodiment

[0058] Figure 2 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a first embodiment of the present invention.

[0059] like Figure 2 As shown, the lithium-ion battery cycle life rapid testing system includes the following components:

[0060] The first analysis unit is used to obtain the internal resistance data, capacity data, and open circuit voltage data of the current lithium-ion battery after it has been tested by radio electromagnetic testing when it is in the factory state, so as to use as multiple radio electromagnetic test data of the current lithium-ion battery.

[0061] Specifically, in the radio electromagnetic testing of lithium-ion batteries, internal resistance testing is a key indicator for measuring the battery's electrical performance. Measured using an AC signal (to avoid polarization interference), it can determine the degree of battery aging. If the internal resistance exceeds 25% of the normal value, it usually means that the capacity has dropped to below 80% of the nominal value.

[0062] Meanwhile, in the radio electromagnetic testing of lithium-ion batteries, capacity testing involves constant current charging and discharging to the cutoff voltage and recording the released amount of electricity. For example, discharging at 25°C can yield a table showing the relationship between temperature and capacity. Generally, the test needs to be repeated three times and the average value taken to improve accuracy.

[0063] The second analysis unit is used to obtain the current lithium-ion battery's volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity as various related parameters of the current lithium-ion battery.

[0064] For example, obtaining the volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity of the current lithium-ion battery as various associated parameters of the current lithium-ion battery includes: using multiple different parameter acquisition units to obtain the volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity of the current lithium-ion battery respectively.

[0065] The third analysis unit is used to take the current lithium-ion battery production batch as the target production batch, and obtain the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch and the interval between each previous batch. The number of lithium-ion batteries produced in each production batch is equal.

[0066] Specifically, lithium-ion battery manufacturers produce batches of lithium-ion batteries with the same model, and the intervals between batches may be the same or different.

[0067] An object assembly mechanism is used to perform multiple trainings on a feedforward neural network to obtain a feedforward neural network after multiple trainings and output it as a cycle lifetime intelligent identification model.

[0068] For example, performing multiple trainings on a feedforward neural network to obtain a multi-trained feedforward neural network and outputting it as a cycle lifetime intelligent assessment model includes: optionally using numerical simulation mode to complete the testing and simulation of the model building process of performing multiple trainings on a feedforward neural network to obtain a multi-trained feedforward neural network and outputting it as a cycle lifetime intelligent assessment model.

[0069] The life assessment mechanism is connected to the first analysis mechanism, the second analysis mechanism, the third analysis mechanism, and the object assembly mechanism, respectively. It is used to intelligently assess the cycle life of the current lithium-ion battery based on multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch, and the interval between each previous batch.

[0070] In this way, by adopting a cycle life numerical analysis mechanism based on an artificial intelligence model to replace the cycle life numerical analysis mechanism that is too biased towards physical measurement and numerical extrapolation, the detection accuracy and detection rate of lithium-ion battery cycle life are improved, while the computational cost, time cost and storage cost of cycle life numerical analysis are reduced, thus meeting the need for rapid one-by-one detection of the cycle life of a large number of lithium-ion batteries.

[0071] The process of training the feedforward neural network multiple times to obtain a trained feedforward neural network and using it as the output of the cycle life intelligent assessment model includes: training the feedforward neural network multiple times to obtain a trained feedforward neural network and using it as the output of the cycle life intelligent assessment model, wherein the number of training times of the feedforward neural network follows the numerical trend of the number of lithium-ion batteries produced in each production batch.

[0072] For example, the training times of the feedforward neural network following the numerical trend of the number of lithium-ion batteries produced in each production batch include: when the number of lithium-ion batteries produced in each production batch is equal and is 10,000, the selected feedforward neural network undergoes 200 training times; when the number of lithium-ion batteries produced in each production batch is equal and is 20,000, the selected feedforward neural network undergoes 400 training times; when the number of lithium-ion batteries produced in each production batch is equal and is 30,000, the selected feedforward neural network undergoes 600 training times, and so on.

[0073] In each training iteration of the feedforward neural network, the known cycle life of a lithium-ion battery that has reached the end of its lifespan is used as the single output of the feedforward neural network. The multiple radio electromagnetic test data of the lithium-ion battery, the various correlation parameters of the lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch before the production batch of the lithium-ion battery are used as the input of the feedforward neural network to complete the training.

[0074] Specifically, the current lithium-ion battery production batch is taken as the target production batch. The average cycle life of each lithium-ion battery in each historical production batch of the same model before the target production batch and the interval between each previous batch are obtained. The number of lithium-ion batteries produced in each production batch is equal. This includes: for each historical production batch, the arithmetic mean of the cycle life of all lithium-ion batteries produced in the historical production batch is the average cycle life of the lithium-ion battery corresponding to the historical production batch, and the interval between the two factory departure times corresponding to the historical production batch and its previous production batch is the interval between the previous batches corresponding to the historical production batch.

[0075] Specifically, for each historical production batch, the arithmetic mean of the cycle life of all lithium-ion batteries produced in the historical production batch is the average cycle life of the lithium-ion batteries corresponding to the historical production batch. The interval between the two production times corresponding to the historical production batch and its previous production batch is the interval between the previous batches corresponding to the historical production batch. The interval between the two production times corresponding to each historical production batch and its previous batch can be unequal, for example, it can be from 5 days to 30 days.

[0076] Among them, the acquisition of internal resistance data, capacity data, and open-circuit voltage data obtained by radio electromagnetic testing of the current lithium-ion battery in its factory state as a multiple radio electromagnetic test data of the current lithium-ion battery includes: acquiring internal resistance data of the current lithium-ion battery in its factory state based on AC signal test mode, and acquiring capacity data of the current lithium-ion battery in its factory state through radio electromagnetic testing by constant current charging and discharging of the current lithium-ion battery to the cutoff voltage to record the released power.

[0077] Second Embodiment

[0078] Figure 3 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a second embodiment of the present invention.

[0079] like Figure 3 As shown, compared to Figure 2 The lithium-ion battery cycle life rapid testing system also includes:

[0080] The marking and judgment mechanism, connected to the life assessment mechanism, is used to receive the current cycle life of the lithium-ion battery and mark the current lithium-ion battery as a defective battery when the current cycle life is lower than the minimum cycle life required to meet the factory requirements, so as to prevent it from entering the market.

[0081] For example, receiving the current cycle life of a lithium-ion battery and marking it as a defective battery to prevent it from entering the market when the current cycle life of a lithium-ion battery is lower than the minimum cycle life required to meet the factory requirements includes: the minimum cycle life required to meet the factory requirements can be selected as 800 cycles;

[0082] The marking and judging mechanism is also used to mark the current lithium-ion battery as a qualified battery so that it can be allowed to enter the market when the current cycle life of the lithium-ion battery is higher than or equal to the minimum cycle life required to meet the factory requirements.

[0083] Third Embodiment

[0084] Figure 4 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a third embodiment of the present invention.

[0085] like Figure 4 As shown, compared to Figure 2 The lithium-ion battery cycle life rapid testing system also includes:

[0086] The test execution mechanism, connected to the first analysis mechanism, is used to perform a radio electromagnetic test on the current lithium-ion battery when a test start signal is received to obtain the internal resistance data, capacity data, and open circuit voltage data of the current lithium-ion battery after resting.

[0087] The test execution mechanism is also used to send the current internal resistance data, capacity data, and open circuit voltage data of the lithium-ion battery after resting to the first analysis mechanism.

[0088] Specifically, the test execution mechanism is also used to send the current internal resistance data, capacity data, and open-circuit voltage data of the lithium-ion battery after resting to the first analysis mechanism, including: the test execution mechanism has a built-in data output interface for sending the current internal resistance data, capacity data, and open-circuit voltage data of the lithium-ion battery after resting to the first analysis mechanism.

[0089] Fourth embodiment

[0090] Figure 5 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a fourth embodiment of the present invention.

[0091] like Figure 5 As shown, compared to Figure 4 The lithium-ion battery cycle life rapid testing system also includes:

[0092] An action triggering mechanism, connected to a test execution mechanism, is used to trigger the test execution mechanism to perform a radio electromagnetic test on the current lithium-ion battery when the current lithium-ion battery is detected to have arrived at the radio electromagnetic test station.

[0093] Specifically, an action triggering mechanism can be composed of an infrared sensing unit and a PLC logic device, which is used to trigger the test execution mechanism to perform a radio electromagnetic test on the current lithium-ion battery when the current lithium-ion battery is detected to have arrived at the radio electromagnetic test station.

[0094] The action triggering mechanism, connected to the test execution mechanism, is used to trigger the test execution mechanism to perform a radio electromagnetic test on the current lithium-ion battery when the current lithium-ion battery is detected to have arrived at the radio electromagnetic test station. This includes: the action triggering mechanism sending a test start signal to the test execution mechanism when it detects that the current lithium-ion battery has arrived at the radio electromagnetic test station.

[0095] Fifth embodiment

[0096] Figure 6 This is an internal structural diagram of a rapid cycle life testing system for lithium-ion batteries according to a fifth embodiment of the present invention.

[0097] like Figure 6 As shown, compared to Figure 2 The lithium-ion battery cycle life rapid testing system also includes:

[0098] The cycle life display mechanism is connected to the life assessment mechanism to receive the current cycle life count of the lithium-ion battery and display the current cycle life count of the lithium-ion battery in real time.

[0099] Specifically, an LCD display array or a liquid crystal display screen can be used to implement the cycle life display mechanism, which is used to receive the current cycle life count of the lithium-ion battery and display the current cycle life count of the lithium-ion battery in real time.

[0100] The cycle count display mechanism, connected to the lifespan assessment mechanism, is used to receive the current cycle life count of the lithium-ion battery and display it in real time. This includes a built-in display cache unit that caches the current cycle life count of the lithium-ion battery after receiving it.

[0101] Next, various embodiments of the present invention will be further described.

[0102] Optionally, within the above embodiments, in the lithium-ion battery cycle life rapid testing system:

[0103] The intelligent cycle life assessment model uses multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch, as well as the interval between each previous batch, to intelligently assess the cycle life of the current lithium-ion battery. This includes simultaneously inputting multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch, as well as the interval between each previous batch, into the intelligent cycle life assessment model.

[0104] Specifically, each historical production batch corresponds to one set of average cycle life of lithium-ion batteries and one set of previous batch interval time.

[0105] The intelligent cycle life assessment model, which uses multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery in each historical production batch of the same model before the target production batch and the interval between each previous batch, also includes: running the intelligent cycle life assessment model to obtain the cycle life of the current lithium-ion battery output by the intelligent cycle life assessment model.

[0106] The process of simultaneously inputting multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch into the cycle life intelligent assessment model includes: performing numerical normalization processing on the multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch before the cycle life intelligent assessment model.

[0107] For example, before synchronously inputting multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch into the cycle life intelligent assessment model, the multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch are subjected to numerical normalization processing, including: the numerical normalization processing is binary numerical conversion processing;

[0108] Among them, the cycle life of the current lithium-ion battery obtained by running the cycle life intelligent assessment model includes: the cycle life of the current lithium-ion battery output by the cycle life intelligent assessment model is a numerical representation after numerical normalization.

[0109] Furthermore, before synchronously inputting the current lithium-ion battery's multiple radio electromagnetic test data, the current lithium-ion battery's various related parameters, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch into the cycle life intelligent assessment model, numerical normalization processing is performed on the current lithium-ion battery's multiple radio electromagnetic test data, the current lithium-ion battery's various related parameters, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch. This includes: selecting different types of programmable logic devices to implement the synchronous input operation and the numerical normalization processing operation respectively.

[0110] And, optionally, in the lithium-ion battery cycle life rapid testing system described above:

[0111] The feedforward neural network is trained multiple times to obtain a trained feedforward neural network, which is then used as the output of the cycle life intelligent assessment model. The numerical trend of the number of training times of the feedforward neural network following the number of lithium-ion batteries produced in each production batch includes: using a parameter transformation function to represent the parameter transformation relationship of the numerical trend of the number of training times of the feedforward neural network following the number of lithium-ion batteries produced in each production batch.

[0112] Specifically, the MATLAB toolbox can be used to simulate and model the representation process of the parameter transformation relationship of the parameter transformation function;

[0113] Among them, the parameter transformation relationship used to represent the trend of the number of training times of the feedforward neural network following the numerical change of the number of lithium-ion batteries produced in each production batch includes: the larger the numerical value of the number of lithium-ion batteries produced in each production batch, the larger the numerical value of the number of training times of the corresponding feedforward neural network.

[0114] Among them, the larger the number of lithium-ion batteries produced in each production batch, the larger the number of training times the corresponding feedforward neural network undergoes. This includes a non-linear positive correlation between the number of lithium-ion batteries produced in each production batch and the number of training times the corresponding feedforward neural network undergoes.

[0115] Furthermore, the parameter transformation relationship, which uses a parameter transformation function to represent the numerical trend of the number of training cycles of the feedforward neural network following the number of lithium-ion batteries produced in each production batch, also includes: in the parameter transformation function, the number of lithium-ion batteries produced in each production batch is the input parameter of the parameter transformation function, and the number of training cycles of the feedforward neural network is the output parameter of the parameter transformation function.

[0116] Sixth Embodiment

[0117] Figure 7 The present invention provides a flowchart of a method for rapid testing of the cycle life of a lithium-ion battery according to a sixth embodiment of the present invention.

[0118] like Figure 7 As shown, the rapid cycle life testing method for lithium-ion batteries includes the following steps:

[0119] Step S71: Obtain the internal resistance data, capacity data, and open circuit voltage data of the current lithium-ion battery after radio electromagnetic testing when it is in the factory state, so as to use as multiple radio electromagnetic test data of the current lithium-ion battery.

[0120] Specifically, in the radio electromagnetic testing of lithium-ion batteries, internal resistance testing is a key indicator for measuring the battery's electrical performance. Measured using an AC signal (to avoid polarization interference), it can determine the degree of battery aging. If the internal resistance exceeds 25% of the normal value, it usually means that the capacity has dropped to below 80% of the nominal value.

[0121] Meanwhile, in the radio electromagnetic testing of lithium-ion batteries, capacity testing involves constant current charging and discharging to the cutoff voltage and recording the released amount of electricity. For example, discharging at 25°C can yield a table showing the relationship between temperature and capacity. Generally, the test needs to be repeated three times and the average value taken to improve accuracy.

[0122] Step S72: Obtain the current lithium-ion battery's volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity as various associated parameters of the current lithium-ion battery;

[0123] For example, obtaining the volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity of the current lithium-ion battery as various associated parameters of the current lithium-ion battery includes: using multiple different parameter acquisition units to obtain the volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity of the current lithium-ion battery respectively.

[0124] Step S73: Take the current lithium-ion battery production batch as the target production batch, obtain the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch and the interval between each previous batch, and ensure that the number of lithium-ion batteries produced in each production batch is equal.

[0125] Specifically, lithium-ion battery manufacturers produce batches of lithium-ion batteries with the same model, and the intervals between batches may be the same or different.

[0126] Step S74: Perform multiple training operations on the feedforward neural network to obtain a feedforward neural network after multiple training operations and output it as the cycle lifetime intelligent identification model.

[0127] For example, performing multiple trainings on a feedforward neural network to obtain a multi-trained feedforward neural network and outputting it as a cycle lifetime intelligent assessment model includes: optionally using numerical simulation mode to complete the testing and simulation of the model building process of performing multiple trainings on a feedforward neural network to obtain a multi-trained feedforward neural network and outputting it as a cycle lifetime intelligent assessment model.

[0128] Step S75: The cycle life intelligent assessment model is used to intelligently assess the cycle life of the current lithium-ion battery based on multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch, and the interval between each previous batch.

[0129] In this way, by adopting a cycle life numerical analysis mechanism based on an artificial intelligence model to replace the cycle life numerical analysis mechanism that is too biased towards physical measurement and numerical extrapolation, the detection accuracy and detection rate of lithium-ion battery cycle life are improved, while the computational cost, time cost and storage cost of cycle life numerical analysis are reduced, thus meeting the need for rapid one-by-one detection of the cycle life of a large number of lithium-ion batteries.

[0130] The process of training the feedforward neural network multiple times to obtain a trained feedforward neural network and using it as the output of the cycle life intelligent assessment model includes: training the feedforward neural network multiple times to obtain a trained feedforward neural network and using it as the output of the cycle life intelligent assessment model, wherein the number of training times of the feedforward neural network follows the numerical trend of the number of lithium-ion batteries produced in each production batch.

[0131] For example, the training times of the feedforward neural network following the numerical trend of the number of lithium-ion batteries produced in each production batch include: when the number of lithium-ion batteries produced in each production batch is equal and is 10,000, the selected feedforward neural network undergoes 200 training times; when the number of lithium-ion batteries produced in each production batch is equal and is 20,000, the selected feedforward neural network undergoes 400 training times; when the number of lithium-ion batteries produced in each production batch is equal and is 30,000, the selected feedforward neural network undergoes 600 training times, and so on.

[0132] In each training iteration of the feedforward neural network, the known cycle life of a lithium-ion battery that has reached the end of its lifespan is used as the single output of the feedforward neural network. The multiple radio electromagnetic test data of the lithium-ion battery, the various correlation parameters of the lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch before the production batch of the lithium-ion battery are used as the input of the feedforward neural network to complete the training.

[0133] Specifically, the current lithium-ion battery production batch is taken as the target production batch. The average cycle life of each lithium-ion battery in each historical production batch of the same model before the target production batch and the interval between each previous batch are obtained. The number of lithium-ion batteries produced in each production batch is equal. This includes: for each historical production batch, the arithmetic mean of the cycle life of all lithium-ion batteries produced in the historical production batch is the average cycle life of the lithium-ion battery corresponding to the historical production batch, and the interval between the two factory departure times corresponding to the historical production batch and its previous production batch is the interval between the previous batches corresponding to the historical production batch.

[0134] Specifically, for each historical production batch, the arithmetic mean of the cycle life of all lithium-ion batteries produced in the historical production batch is the average cycle life of the lithium-ion batteries corresponding to the historical production batch. The interval between the two production times corresponding to the historical production batch and its previous production batch is the interval between the previous batches corresponding to the historical production batch. The interval between the two production times corresponding to each historical production batch and its previous batch can be unequal, for example, it can be from 5 days to 30 days.

[0135] Among them, the acquisition of internal resistance data, capacity data, and open-circuit voltage data obtained by radio electromagnetic testing of the current lithium-ion battery in its factory state as a multiple radio electromagnetic test data of the current lithium-ion battery includes: acquiring internal resistance data of the current lithium-ion battery in its factory state based on AC signal test mode, and acquiring capacity data of the current lithium-ion battery in its factory state through radio electromagnetic testing by constant current charging and discharging of the current lithium-ion battery to the cutoff voltage to record the released power.

[0136] Furthermore, in a rapid lithium-ion battery cycle life testing system and method according to the present invention:

[0137] The method of performing multiple trainings on a feedforward neural network to obtain a feedforward neural network after multiple trainings and using it as the output of a cycle lifetime intelligent identification model also includes: the feedforward neural network consists of an output layer, an input layer and multiple hidden layers, wherein the multiple hidden layers are located between the output layer and the input layer;

[0138] The feedforward neural network consists of an output layer, an input layer, and multiple hidden layers. The multiple hidden layers are located between the output layer and the input layer. In the feedforward neural network, the number of hidden layers follows the numerical trend of the batch number of each historical production batch of the same model before the target production batch.

[0139] For example, in the feedforward neural network, the numerical trend of the number of hidden layers following the batch number of each historical production batch of the same model before the target production batch includes: using a numerical mapping formula to represent the numerical mapping relationship of the numerical trend of the number of hidden layers following the batch number of each historical production batch of the same model before the target production batch.

[0140] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of embodiments of this application, "multiple" means two or more, unless otherwise explicitly specified. In the description of this specification, references to terms such as "an embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of this application.

[0141] In this specification, the illustrative expressions of the terms used do not necessarily refer to the same implementation or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more implementations or examples. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A rapid cycle life testing system for lithium-ion batteries, characterized in that, The system includes: The first analysis unit is used to obtain the internal resistance data, capacity data, and open circuit voltage data of the current lithium-ion battery after it has been tested by radio electromagnetic testing when it is in the factory state, so as to use as multiple radio electromagnetic test data of the current lithium-ion battery. The second analysis unit is used to obtain the current lithium-ion battery's volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity as various related parameters of the current lithium-ion battery. The third analysis unit is used to take the current lithium-ion battery production batch as the target production batch, and obtain the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch and the interval between each previous batch. The number of lithium-ion batteries produced in each production batch is equal. An object assembly mechanism is used to perform multiple trainings on a feedforward neural network to obtain a feedforward neural network after multiple trainings and output it as a cycle lifetime intelligent identification model. The lifespan assessment mechanism is connected to the first analysis mechanism, the second analysis mechanism, the third analysis mechanism, and the object assembly mechanism, respectively. It is used to intelligently assess the cycle life of the current lithium-ion battery using a cycle life intelligent assessment model based on multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch, as well as the interval between each previous batch.

2. The lithium-ion battery cycle life rapid testing system as described in claim 1, characterized in that: The process of training the feedforward neural network multiple times to obtain a trained feedforward neural network and using it as the output of the cycle life intelligent assessment model includes: training the feedforward neural network multiple times to obtain a trained feedforward neural network and using it as the output of the cycle life intelligent assessment model, wherein the number of training times of the feedforward neural network follows the numerical trend of the number of lithium-ion batteries produced in each production batch. In each training iteration of the feedforward neural network, the known cycle life of a lithium-ion battery that has reached the end of its lifespan is used as the single output of the feedforward neural network. The multiple radio electromagnetic test data of the lithium-ion battery, the various correlation parameters of the lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch before the production batch of the lithium-ion battery are used as the input of the feedforward neural network to complete the training.

3. The lithium-ion battery cycle life rapid testing system as described in claim 2, characterized in that: The current lithium-ion battery production batch is taken as the target production batch. The average cycle life of each lithium-ion battery and the interval between each previous batch are obtained for each historical production batch of the same model before the target production batch. The number of lithium-ion batteries produced in each production batch is equal. For each historical production batch, the arithmetic mean of the cycle life of all lithium-ion batteries produced in the historical production batch is the average cycle life of the lithium-ion battery corresponding to the historical production batch. The interval between the two factory departure times corresponding to the historical production batch and its previous production batch is the interval between the previous batch corresponding to the historical production batch. The acquisition of internal resistance data, capacity data, and open-circuit voltage data obtained from radio electromagnetic testing of the current lithium-ion battery at the time of delivery, as well as the data obtained after resting, as multiple radio electromagnetic test data of the current lithium-ion battery, includes: acquiring internal resistance data of the current lithium-ion battery at the time of delivery based on AC signal test mode, and acquiring capacity data of the current lithium-ion battery at the time of delivery by constant current charging and discharging to the cutoff voltage to record the released power.

4. The rapid cycle life testing system for lithium-ion batteries as described in claim 3, characterized in that, The system also includes: The marking and judgment mechanism, connected to the life assessment mechanism, is used to receive the current cycle life of the lithium-ion battery and mark the current lithium-ion battery as a defective battery when the current cycle life is lower than the minimum cycle life required to meet the factory requirements, so as to prevent it from entering the market. The marking and judging mechanism is also used to mark the current lithium-ion battery as a qualified battery so that it can be allowed to enter the market when the current cycle life of the lithium-ion battery is higher than or equal to the minimum cycle life required to meet the factory requirements.

5. The rapid cycle life testing system for lithium-ion batteries as described in claim 3, characterized in that, The system also includes: The test execution mechanism, connected to the first analysis mechanism, is used to perform a radio electromagnetic test on the current lithium-ion battery when a test start signal is received to obtain the internal resistance data, capacity data, and open circuit voltage data of the current lithium-ion battery after resting. The test execution mechanism is also used to send the current internal resistance data, capacity data, and open-circuit voltage data of the lithium-ion battery after resting to the first analysis mechanism.

6. The rapid cycle life testing system for lithium-ion batteries as described in claim 5, characterized in that, The system also includes: An action triggering mechanism, connected to a test execution mechanism, is used to trigger the test execution mechanism to perform a radio electromagnetic test on the current lithium-ion battery when the current lithium-ion battery is detected to have arrived at the radio electromagnetic test station. The action triggering mechanism, connected to the test execution mechanism, is used to trigger the test execution mechanism to perform a radio electromagnetic test on the current lithium-ion battery when the current lithium-ion battery is detected to have arrived at the radio electromagnetic test station. This includes: the action triggering mechanism sending a test start signal to the test execution mechanism when it detects that the current lithium-ion battery has arrived at the radio electromagnetic test station.

7. The rapid cycle life testing system for lithium-ion batteries as described in claim 3, characterized in that, The system also includes: The cycle life display mechanism is connected to the life assessment mechanism to receive the current cycle life count of the lithium-ion battery and display the current cycle life count of the lithium-ion battery in real time. The cycle count display mechanism, connected to the lifespan assessment mechanism, is used to receive the current cycle life count of the lithium-ion battery and display it in real time. This includes a built-in display cache unit that caches the current cycle life count of the lithium-ion battery after receiving it.

8. The rapid cycle life testing system for lithium-ion batteries as described in any one of claims 3-7, characterized in that: The intelligent cycle life assessment model uses multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch, as well as the interval between each previous batch, to intelligently assess the cycle life of the current lithium-ion battery. This includes simultaneously inputting multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery corresponding to each historical production batch of the same model before the target production batch, as well as the interval between each previous batch, into the intelligent cycle life assessment model. The intelligent cycle life assessment model, which uses multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life of each lithium-ion battery in each historical production batch of the same model before the target production batch and the interval between each previous batch, also includes: running the intelligent cycle life assessment model to obtain the cycle life of the current lithium-ion battery output by the intelligent cycle life assessment model. The process of simultaneously inputting multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch into the cycle life intelligent assessment model includes: performing numerical normalization processing on the multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch before the cycle life intelligent assessment model. Among them, the cycle life of the current lithium-ion battery obtained by running the cycle life intelligent assessment model includes: the cycle life of the current lithium-ion battery output by the cycle life intelligent assessment model is a numerical representation after numerical normalization. Before synchronously inputting the current lithium-ion battery's multiple radio electromagnetic test data, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch into the cycle life intelligent assessment model, numerical normalization processing is performed on the current lithium-ion battery's multiple radio electromagnetic test data, various related parameters of the current lithium-ion battery, and the average cycle life and the interval between previous batches of lithium-ion batteries corresponding to each historical production batch of the same model before the target production batch. This includes selecting different types of programmable logic devices to implement the synchronous input operation and numerical normalization processing operation respectively.

9. The rapid cycle life testing system for lithium-ion batteries as described in any one of claims 3-7, characterized in that: The feedforward neural network is trained multiple times to obtain a trained feedforward neural network, which is then used as the output of the cycle life intelligent assessment model. The numerical trend of the number of training times of the feedforward neural network following the number of lithium-ion batteries produced in each production batch includes: using a parameter transformation function to represent the parameter transformation relationship of the numerical trend of the number of training times of the feedforward neural network following the number of lithium-ion batteries produced in each production batch. Among them, the parameter transformation relationship used to represent the trend of the number of training times of the feedforward neural network following the numerical change of the number of lithium-ion batteries produced in each production batch includes: the larger the numerical value of the number of lithium-ion batteries produced in each production batch, the larger the numerical value of the number of training times of the corresponding feedforward neural network. Among them, the larger the number of lithium-ion batteries produced in each production batch, the larger the number of training times the corresponding feedforward neural network undergoes. This includes a non-linear positive correlation between the number of lithium-ion batteries produced in each production batch and the number of training times the corresponding feedforward neural network undergoes. The parameter transformation relationship, which uses a parameter transformation function to represent the numerical trend of the number of training cycles of the feedforward neural network following the number of lithium-ion batteries produced in each production batch, further includes: in the parameter transformation function, the number of lithium-ion batteries produced in each production batch is the input parameter of the parameter transformation function, and the number of training cycles of the feedforward neural network is the output parameter of the parameter transformation function.

10. A method for rapid detection of the cycle life of a lithium-ion battery, characterized in that, The method includes: The internal resistance data, capacity data, and open-circuit voltage data obtained from the radio electromagnetic testing of the current lithium-ion battery at the factory state are used as multiple radio electromagnetic test data of the current lithium-ion battery. Obtain the current lithium-ion battery's volume, weight, minimum operating temperature, maximum operating temperature, minimum operating humidity, and maximum operating humidity as various related parameters of the current lithium-ion battery; The current lithium-ion battery production batch is taken as the target production batch. The average cycle life of each lithium-ion battery and the interval between each previous batch are obtained for each historical production batch of the same model before the target production batch. The number of lithium-ion batteries produced in each production batch is equal. The feedforward neural network is trained multiple times to obtain a feedforward neural network after multiple trainings, which is then used as the output of the cycle lifetime intelligent identification model. The cycle life intelligent assessment model uses multiple radio electromagnetic test data of the current lithium-ion battery, various related parameters of the current lithium-ion battery, the average cycle life of each lithium-ion battery in each historical production batch of the same model before the target production batch, and the interval between each previous batch to intelligently assess the cycle life of the current lithium-ion battery.

Images