Lithium battery state estimation system based on big data

Abstract

The invention discloses a lithium battery state estimation system based on big data. The system comprises a data acquisition module used for acquiring real-time operation data of a lithium battery and obtaining offline aging test data of the lithium battery; the method relates to the technical field of battery health state evaluation, an equivalent circuit model is constructed through historical operation data of a lithium battery and offline aging test data of the lithium battery, the equivalent circuit model retains the interpretability of a circuit model, and the lithium battery health state is evaluated by using equivalent circuit parameters obtained by the equivalent circuit model. The state estimation model is constructed through the long-short-term memory network algorithm, the real-time calculation precision of the state estimation model for calculating the SOC and SOH of the lithium battery is improved by using the time sequence characteristic capture capability of the long-short-term memory network algorithm, the offline aging data is fully utilized in the process, the aging trend capture capability of the state estimation model is enhanced, and the real-time calculation precision of the state estimation model for calculating the SOC and SOH of the lithium battery is improved. Therefore, the estimation precision of the state estimation model on the SOC and SOH of the lithium battery is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of battery health state assessment, and particularly to a lithium battery state prediction system based on big data. Background Art

[0002] As an efficient energy storage device, lithium batteries have been widely used in fields such as electric vehicles, portable electronic devices, and large-scale energy storage systems. The quality of their performance directly affects the performance and safety of related devices. Therefore, accurate prediction of the state of lithium batteries is crucial.

[0003] Publication No. CN112034349B discloses an online prediction method for the health state of lithium batteries, and publication No. CN113075560B discloses an online prediction method for the health state of power lithium-ion batteries. However, the above applications still have the following problems: CN112034349B performs SOH prediction through the Thevenin model combined with the extended Kalman filter, and CN113075560B predicts SOH based on the second-order RC model and the BP neural network. Both CN112034349B and CN113075560B rely on a single equivalent circuit model and do not make full use of offline aging data, resulting in insufficient ability of the model to capture the aging trend and thus a decrease in long-term prediction accuracy. Summary of the Invention

[0004] To solve the technical problems in the background art, the present invention proposes a lithium battery state prediction system based on big data.

[0005] The lithium battery state prediction system based on big data proposed by the present invention includes: Data acquisition module: used to collect the real-time operation data of the lithium battery and obtain the offline aging test data of the lithium battery; The operation data includes terminal voltage, charge and discharge current, temperature, number of cycles, and historical SOH; The offline aging test data refers to the data collected through systematic aging experiments on the lithium battery in a laboratory environment; Data preprocessing module: used to preprocess the data collected by the data acquisition module; State prediction module: used to calculate the SOC and SOH of the lithium battery in real time by obtaining the historical operation data of the lithium battery and the offline aging test data of the lithium battery; Preferably, the state prediction module includes: State prediction model construction unit: Obtain the historical operation data of the lithium battery and the offline aging test data of the lithium battery, and construct an equivalent circuit model of the lithium battery through the Thevenin model: Based on the historical operation data of the lithium battery, the offline aging test data of the lithium battery, and the equivalent circuit parameters obtained from the equivalent circuit model, a state prediction model is constructed through the long short-term memory network algorithm. The state prediction model is used to calculate the SOC and SOH of the lithium battery; The equivalent circuit parameters include ohmic internal resistance, polarization internal resistance, and polarization capacitance; State prediction output unit: Through the data preprocessing module, the real-time operation data of the lithium battery, the offline aging test data of the lithium battery, and the equivalent circuit parameters obtained from the equivalent circuit model are obtained. The state prediction model calculates the SOC and SOH of the lithium battery in real time.

[0006] Preferably, the state prediction module further includes: State prediction correction unit: Obtain the true values of the SOC and SOH of the lithium battery through offline testing, and calculate the error values of the SOC and SOH of the lithium battery; Set the error threshold of SOC or SOH. When the error value of SOC or SOH is less than the error threshold, perform coarse-grained correction on the SOC and SOH obtained through the state prediction model; The core purpose of coarse-grained correction is to quickly process low-level errors through a simplified strategy while ensuring the correction accuracy, reduce the computational complexity, and improve the real-time performance of the system; When the error value of SOC or SOH is greater than or equal to the error threshold, use the fuzzy logic algorithm to correct the SOC and SOH obtained through the state prediction model.

[0007] Preferably, in the state prediction correction unit, when setting the error threshold of SOC or SOH, perform real-time calculations of the SOC and SOH of the lithium battery n times through the state prediction model, and obtain the corresponding true values of the SOC and SOH of the lithium battery n times through offline testing, and calculate the error values of the SOC and SOH of the lithium battery n times; The error threshold of SOC or SOH = , where μ is the mean of the error values of SOC or SOH n times, σ is the standard deviation of the error values of SOC or SOH n times, is the scaling factor, 0 < .

[0008] Preferably, in the state prediction correction unit, the scaling factor , where μ is the mean of the error values of SOC or SOH n times, σ is the standard deviation of the error values of SOC or SOH n times, α is the weight coefficient, 0 < α < 1, and SOH is obtained through the state prediction model, is the preset critical value of SOH, and exp is the exponential function in mathematics.

[0009] Preferably, it further includes: Remaining useful life prediction module: Obtain the offline aging test data of the lithium battery through the data preprocessing module and the SOH obtained by the state estimation module, and predict the remaining useful life of the lithium battery through the recurrent neural network algorithm.

[0010] Preferably, in the data preprocessing module, the preprocessing includes: Clean, standardize, and extract features from the real-time operation data of the lithium battery collected by the data acquisition module and the offline aging test data of the lithium battery obtained. Among them, the feature extraction is: Extract the characteristic parameters related to the aging of the lithium battery. The characteristic parameters related to the aging of the lithium battery include polarization voltage, ohmic internal resistance, polarization internal resistance, and capacitance parameters.

[0011] A method for estimating the state of a lithium battery based on big data includes the following steps: S1. Collect the operation data of the lithium battery in real time and obtain the offline aging test data of the lithium battery, and preprocess the operation data of the lithium battery and the offline aging test data of the lithium battery; S2. Obtain the historical operation data of the lithium battery and the offline aging test data of the lithium battery, and construct an equivalent circuit model of the lithium battery through the Thevenin model: Based on the historical operation data of the lithium battery, the offline aging test data of the lithium battery, and the equivalent circuit parameters obtained from the equivalent circuit model, construct a state estimation model through the long short-term memory network algorithm. The state estimation model is used to calculate the SOC and SOH of the lithium battery; S3. Obtain the real-time operation data of the lithium battery in S1, the offline aging test data of the lithium battery, and the equivalent circuit parameters obtained from the equivalent circuit model in S2, and calculate the SOC and SOH of the lithium battery in real time through the state estimation model; S4. Obtain the true values of the SOC and SOH of the lithium battery through offline testing, and calculate the error values of the SOC and SOH of the lithium battery; Set the error threshold of the SOC or SOH. When the error value of the SOC or SOH is less than the error threshold, perform coarse-grained correction on the SOC and SOH obtained through the state estimation model; When the error value of the SOC or SOH is greater than or equal to the error threshold, use the fuzzy logic algorithm to correct the SOC and SOH obtained through the state estimation model; S5. Through the offline aging test data of the lithium battery obtained in S1 and the SOH calculated in real time by the state estimation model in S3, predict the remaining useful life of the lithium battery through the recurrent neural network algorithm.

[0012] In the present invention, the proposed lithium battery state estimation system based on big data has the following beneficial technical effects: Further Explanation of Inventiveness In addition to the inventive points mentioned above, the present invention further demonstrates innovation in the following aspects: 1. By using the historical operation data of the lithium battery and the off-line aging test data of the lithium battery, an equivalent circuit model is constructed. The equivalent circuit model retains the interpretability of the circuit model, and the equivalent circuit parameters obtained from the equivalent circuit model are used to construct a state prediction model through the long short-term memory network algorithm. By utilizing the time series feature capture ability of the long short-term memory network algorithm, the real-time calculation accuracy of calculating the SOC and SOH of the lithium battery by the state prediction model is improved. Moreover, the off-line aging data is fully utilized during the process to enhance the capture ability of the state prediction model for the aging trend, thereby further improving the prediction accuracy of the state prediction model for the SOC and SOH of the lithium battery.

[0013] 2. By setting up the state prediction module, while realizing the real-time calculation of the SOC and SOH of the lithium battery, by setting the error threshold of the SOC or SOH, in the state prediction correction unit, when the error value of the SOC or SOH is less than the error threshold, the SOC and SOH obtained through the state prediction model are coarsely corrected to improve the correction efficiency; when the error value of the SOC or SOH is greater than or equal to the error threshold, the fuzzy logic algorithm is used to correct the SOC and SOH obtained through the state prediction model to ensure the accuracy of the correction. While ensuring the accuracy, the calculation complexity is reduced, and different degrees of prediction errors can be more flexibly handled.

[0014] 3. When setting the error threshold of the SOC or SOH, the SOC and SOH of the lithium battery obtained by performing n times of real-time calculations of the state prediction model are used to quantify the error distribution range, avoiding the subjectivity of the traditional fixed threshold. Through the weighted fusion of the error statistical term and the SOH-related term, the calculation of the scaling factor k is realized. k is adjusted in real time according to the SOH to ensure that the threshold matches the battery aging degree, making the error threshold of the SOC or SOH more in line with the actual state of the battery, realizing the adaptive division of the error threshold of the SOC or SOH, and enhancing the self-adaptability of the error correction.

[0015] The additional aspects and advantages of the present invention will be partially given in the following description, partially will become obvious from the following description, or will be understood through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is the principle block diagram of the system of the present invention; Figure 2 is the flowchart of the method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0017] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation of the present invention.

[0018] As Figure 1 shown, a lithium battery state prediction system based on big data includes: Data acquisition module: used to collect real-time operation data of the lithium battery and obtain offline aging test data of the lithium battery; The operation data includes terminal voltage, charge and discharge current, temperature, number of cycles, and historical SOH; The offline aging test data refers to the data collected through systematic aging experiments on the lithium battery in a laboratory environment; Data preprocessing module: used to preprocess the data collected by the data acquisition module; Further, in the data preprocessing module, the preprocessing includes: Cleaning, standardizing, and feature extraction are performed on the real-time operation data of the lithium battery collected by the data acquisition module and the offline aging test data of the lithium battery obtained. Among them, the feature extraction is: Extract feature parameters related to lithium battery aging. The feature parameters related to lithium battery aging include polarization voltage, ohmic internal resistance, polarization internal resistance, and capacitance parameters.

[0019] By comprehensively collecting the real-time operation data and offline aging test data of the lithium battery, the aging characteristics of polarization voltage, ohmic internal resistance, polarization internal resistance, and capacitance parameters are extracted through data preprocessing. It breaks through the limitations of a single data type and effectively improves the comprehensive characterization ability of the battery aging process.

[0020] State prediction module: used to calculate the SOC and SOH of the lithium battery in real time by obtaining the historical operation data of the lithium battery and the offline aging test data of the lithium battery; Further, the state prediction module includes: State prediction model construction unit: Obtain the historical operation data of the lithium battery and the offline aging test data of the lithium battery, and construct an equivalent circuit model of the lithium battery through the Thevenin model: According to the equivalent circuit parameters obtained from the historical operation data of the lithium battery, the offline aging test data of the lithium battery, and the equivalent circuit model, construct a state prediction model through the long short-term memory network algorithm. The state prediction model is used to calculate the SOC and SOH of the lithium battery; The equivalent circuit parameters include ohmic internal resistance, polarization internal resistance, and polarization capacitance; Thevenin model is an analytical tool in circuit theory, which is used to convert a complex linear active two-terminal network into a simplified circuit model. Its core idea is to simplify any linear network containing power supply and resistance into a series combination of an ideal voltage source and a resistor through equivalent transformation.

[0021] SOC is the state of charge of the lithium battery, and SOH is the percentage of the current available capacity of the lithium battery to the rated capacity; State estimation output unit: The real-time operation data of the lithium battery, the offline aging test data of the lithium battery and the equivalent circuit parameters obtained by the equivalent circuit model are obtained through the data preprocessing module, and the state estimation model calculates the SOC and SOH of the lithium battery in real time.

[0022] An equivalent circuit model is established based on the Thevenin model, and a state prediction model is constructed by combining the long short-term memory network algorithm. The interpretability of the physical model and the nonlinear fitting ability of deep learning are integrated to improve the real-time calculation accuracy of SOC and SOH.

[0023] An equivalent circuit model is constructed through the historical operation data of lithium batteries and the offline aging test data of lithium batteries. The equivalent circuit model retains the interpretability of the circuit model, and the equivalent circuit parameters obtained by the equivalent circuit model are used to construct a state prediction model through the long short-term memory network algorithm. The timing feature capture capability of the long short-term memory network algorithm is utilized to improve the real-time calculation accuracy of the state prediction model for calculating the SOC and SOH of lithium batteries. The process also makes full use of offline aging data to enhance the state prediction model's ability to capture aging trends, thereby further improving the state prediction model's estimation accuracy for the SOC and SOH of lithium batteries.

[0024] Furthermore, the state estimation module also includes: State estimation correction unit: obtains the true value of the lithium battery SOC and SOH through offline testing, and obtains the error value of the lithium battery SOC and SOH through calculation; The error values ​​of SOC and SOH can be obtained by calculating the difference between the actual values ​​of SOC and SOH and the SOC and SOH obtained by the state estimation model; Setting an error threshold of SOC or SOH. When the error value of SOC or SOH is less than the error threshold, a coarse-grained correction is performed on the SOC and SOH obtained by the state estimation model. Coarse-grained correction refers to correcting model bias or compensating for system errors through larger adjustments or periodic global updates; The core purpose of coarse-grained correction is to quickly process low-level errors through simplified strategies while ensuring the accuracy of correction, thereby reducing computational complexity and improving system real-time performance. When the error values of SOC or SOH are greater than or equal to the error threshold, the fuzzy logic algorithm is used to correct the SOC and SOH obtained through the state prediction model.

[0025] Through the setting of the state prediction module, while realizing the real-time calculation of the SOC and SOH of the lithium battery, by setting the error threshold of SOC or SOH, in the state prediction correction unit, when the error values of SOC or SOH are less than the error threshold, the SOC and SOH obtained through the state prediction model are coarsely corrected to improve the correction efficiency; when the error values of SOC or SOH are greater than or equal to the error threshold, the fuzzy logic algorithm is used to correct the SOC and SOH obtained through the state prediction model to ensure the accuracy of the correction, reduce the computational complexity on the premise of ensuring the accuracy, and more flexibly cope with different degrees of prediction errors.

[0026] Furthermore, in the state prediction correction unit, when setting the error threshold of SOC or SOH, the SOC and SOH of the lithium battery obtained by performing n real-time calculations of the state prediction model are obtained, and the corresponding true values of the SOC and SOH of the lithium battery for n times are obtained through off-line testing, and the error values of the SOC and SOH of the lithium battery for n times are calculated; n is a positive integer greater than 3; The SOC and SOH of the lithium battery obtained by performing n real-time calculations of the state prediction model can be obtained by setting n samples of lithium batteries, and the SOC and SOH of each lithium battery sample are obtained through the real-time calculation of the state prediction model, so as to obtain the true values of the SOC and SOH of n lithium batteries; The error threshold of SOC or SOH = , where μ is the mean value of the error values of SOC or SOH for n times, σ is the standard deviation of the error values of SOC or SOH for n times, is a scaling factor, 0 < .

[0027] Furthermore, in the state prediction correction unit, the scaling factor k is: , in the formula, the denominator is not 0, μ is the mean value of the error values of SOC or SOH for n times, σ is the standard deviation of the error values of SOC or SOH for n times, α is a weight coefficient, 0 < α < 1, and SOH is obtained through the state prediction model, is the preset critical value of SOH. In an optional embodiment, = 80%, and exp is the exponential function in mathematics; Through the weighted fusion of the error statistical terms μ, σ and the SOH-related terms SOH, the calculation of the scaling factor k is realized, and the influence of working condition fluctuations and battery aging on the threshold is balanced.

[0028] α is used to balance the contributions of the error statistical term and the SOH-related term to k; The SOC and SOH of the lithium battery are obtained by performing real-time calculations of the state prediction model n times. The mean value μ and standard deviation σ of the errors of SOC and SOH are calculated to quantify the error distribution range. Compared with the subjectivity of the traditional fixed threshold, this method is based on the statistical law of measured data to define the reasonable fluctuation range of errors, improving the objectivity and credibility of threshold setting; Introduce the scaling factor formula , balance the error fluctuations μ, σ and the contribution of SOH to the error threshold of SOC or SOH. Through the exponential function Realize the non-linear response of the threshold to the aging degree, enable the error threshold of SOC or SOH to dynamically adapt to the working condition fluctuations and the aging process, avoid the limitations of a single dimension, and k is adjusted in real time according to SOH to ensure that the threshold matches the battery aging degree, making the error threshold of SOC or SOH more in line with the actual state of the battery, realizing the adaptive division of the SOC or SOH error threshold and improving the self-adaptability of error correction.

[0029] Furthermore, it also includes: Remaining life prediction module: Obtain the offline aging test data of the lithium battery and the SOH obtained by the state prediction module through the data preprocessing module, and predict the remaining service life of the lithium battery through the recurrent neural network algorithm, providing a scientific basis for subsequent battery maintenance and replacement.

[0030] Such as Figure 2 A method for predicting the state of a lithium battery based on big data shown, including the following steps: S1. Real-time collect the operation data of the lithium battery and obtain the offline aging test data of the lithium battery, and preprocess the operation data of the lithium battery and the offline aging test data of the lithium battery; S2. Obtain the historical operation data of the lithium battery and the offline aging test data of the lithium battery, and construct an equivalent circuit model of the lithium battery through the Thevenin model: According to the equivalent circuit parameters obtained from the historical operation data of the lithium battery, the offline aging test data of the lithium battery and the equivalent circuit model, construct a state prediction model through the long short-term memory network algorithm, and the state prediction model is used to calculate the SOC and SOH of the lithium battery; S3. Obtain the real-time operation data of the lithium battery in S1, the offline aging test data of the lithium battery and the equivalent circuit parameters obtained from the equivalent circuit model in S2, and calculate the SOC and SOH of the lithium battery in real time through the state prediction model; S4. Obtain the true values of SOC and SOH of the lithium battery through offline testing, and calculate the error values of SOC and SOH of the lithium battery; Set the error threshold for SOC or SOH. When the error value of SOC or SOH is less than the error threshold, perform a coarse-grained correction on the SOC and SOH obtained through the state prediction model. When the error value of SOC or SOH is greater than or equal to the error threshold, use the fuzzy logic algorithm to correct the SOC and SOH obtained through the state prediction model. S5. Through the offline aging test data of the lithium battery obtained in S1 and the SOH calculated in real time by the state prediction model in S3, predict the remaining service life of the lithium battery through the recurrent neural network algorithm.

[0031] In summary, the present application realizes the prediction of the state and life of the lithium battery, providing strong support for the safe and reliable operation of new energy equipment.

[0032] At the same time, the content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

[0033] In the embodiments provided by the present invention, it should be understood that the disclosed system or method can be implemented in other ways. For example, the above-described embodiments of the invention are merely illustrative. For example, the division of modules is only a logical function division, and there may be other division methods in actual implementation.

[0034] The modules described as separate components may or may not be physically separated. The components shown as modules may or may not be physical modules. They may be located in one place or distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0035] In addition, in each embodiment of the present invention, the functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of a combination of hardware and software functional modules.

[0036] For those skilled in the field of operation and maintenance, it is obvious that the present invention is not limited to the details of the above-described exemplary embodiments, and can be implemented in other specific forms without departing from the basic characteristics of the present invention.

[0037] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A lithium battery status prediction system based on big data, characterized in that: include: Data acquisition module: used to collect real-time operation data of lithium batteries and obtain offline aging test data of lithium batteries; Data preprocessing module: used to preprocess the data collected by the data acquisition module; State estimation module: It is used to calculate the SOC and SOH of lithium batteries in real time by obtaining the historical operation data of lithium batteries and offline aging test data of lithium batteries.

2. The lithium battery state prediction system based on big data according to claim 1 is characterized in that: The state estimation module includes: State prediction model building unit: Obtain the historical operation data of the lithium battery and the offline aging test data of the lithium battery, and build the equivalent circuit model of the lithium battery through the Thevenin model: Based on the historical operation data of lithium batteries, offline aging test data of lithium batteries and equivalent circuit parameters obtained from the equivalent circuit model, a state prediction model is constructed through the long short-term memory network algorithm. The state prediction model is used to calculate the SOC and SOH of lithium batteries. State estimation output unit: The real-time operation data of the lithium battery, the offline aging test data of the lithium battery and the equivalent circuit parameters obtained by the equivalent circuit model are obtained through the data preprocessing module, and the state estimation model calculates the SOC and SOH of the lithium battery in real time.

3. The lithium battery state prediction system based on big data according to claim 2 is characterized in that: The state estimation module also includes: State estimation correction unit: obtains the true value of the lithium battery SOC and SOH through offline testing, and obtains the error value of the lithium battery SOC and SOH through calculation; Set an error threshold of SOC or SOH. When the error value of SOC or SOH is less than the error threshold, perform a coarse-grained correction on the SOC and SOH obtained by the state estimation model. When the error value of SOC or SOH is greater than or equal to the error threshold, the fuzzy logic algorithm is used to correct the SOC and SOH obtained by the state estimation model.

4. The lithium battery state prediction system based on big data according to claim 3 is characterized in that: In the state estimation correction unit, when setting the SOC or SOH error threshold, the SOC and SOH of the lithium battery are calculated in real time by the state estimation model n times, and the corresponding n times of the real value of the SOC and SOH of the lithium battery are obtained through offline testing, and the error values ​​of the SOC and SOH of the lithium battery n times are calculated; Error threshold of SOC or SOH = , where μ is the mean of the n-times SOC or SOH error values, σ is the standard deviation of the n-times SOC or SOH error values, is the scaling factor, 0< .

5. The lithium battery state prediction system based on big data according to claim 4 is characterized in that: In the state estimation correction unit, the scaling factor , where α is the weight coefficient, 0<α<1, SOH is obtained through the state prediction model, is the preset critical value of SOH, and exp is the exponential function in mathematics.

6. The lithium battery state prediction system based on big data according to claim 1 is characterized in that: Also includes: Remaining life prediction module: The offline aging test data of the lithium battery is obtained through the data preprocessing module and the SOH obtained by the state estimation module, and the remaining service life of the lithium battery is predicted through the recurrent neural network algorithm.

7. The lithium battery state prediction system based on big data according to claim 1 is characterized in that: In the data preprocessing module, preprocessing includes: The data acquisition module collects real-time operation data of lithium batteries and obtains offline aging test data of lithium batteries, and performs cleaning, standardization and feature extraction, wherein the feature extraction is: The characteristic parameters related to lithium battery aging are extracted. The characteristic parameters related to lithium battery aging include polarization voltage, ohmic internal resistance, polarization internal resistance and capacitance parameters.

8. The method for estimating lithium battery status based on big data according to any one of claims 1 to 7, characterized in that: The following steps are involved: S1. Collecting the operating data of the lithium battery in real time and obtaining the offline aging test data of the lithium battery, and preprocessing the operating data of the lithium battery and the offline aging test data of the lithium battery; S2. Obtain the historical operation data of the lithium battery and the offline aging test data of the lithium battery, and construct the equivalent circuit model of the lithium battery through the Thevenin model: Based on the historical operation data of lithium batteries, offline aging test data of lithium batteries and equivalent circuit parameters obtained from the equivalent circuit model, a state prediction model is constructed through the long short-term memory network algorithm. The state prediction model is used to calculate the SOC and SOH of lithium batteries. S3, obtaining the real-time operation data of the lithium battery in S1, the offline aging test data of the lithium battery and the equivalent circuit parameters obtained by the equivalent circuit model in S2, and calculating the SOC and SOH of the lithium battery in real time through the state prediction model; S4. Obtain the true values ​​of the SOC and SOH of the lithium battery through offline testing, and obtain the error values ​​of the SOC and SOH of the lithium battery through calculation; Set an error threshold of SOC or SOH. When the error value of SOC or SOH is less than the error threshold, perform a coarse-grained correction on the SOC and SOH obtained by the state estimation model. When the error value of SOC or SOH is greater than or equal to the error threshold, the fuzzy logic algorithm is used to correct the SOC and SOH obtained by the state estimation model; S5, using the offline aging test data of the lithium battery obtained in S1 and the SOH calculated in real time by the state prediction model in S3, predicts the remaining service life of the lithium battery through a recurrent neural network algorithm.

Images