Satellite-borne lithium ion battery life prediction method based on multi-working-condition degradation data

By constructing GBM process model and migration model, using multi-condition degradation data to predict the life of lithium-ion batteries, the problem of low degradation equivalent modeling efficiency and insufficient samples under low stress conditions is solved, and high precision and fast life is achieved predict.

CN118965740APending Publication Date: 2024-11-15HARBIN INST OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411010310.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

In the prior art, the equivalent modeling efficiency of lithium-ion batteries is low, the remaining life prediction requires a long period, and the life prediction cannot be completed when effective samples are lacking under low stress conditions.

Method used

The lithium-ion battery degradation equivalent model is constructed based on the GBM process, the likelihood function is obtained, and the satellite-borne lithium-ion battery life prediction is combined with the migration model. The multi-condition degradation data is used for linear migration, and an efficient basic model for life prediction is established.

Benefits of technology

It realizes simplification of battery degradation modeling, shortens life prediction time, can complete life prediction of large number of battery models in a short time, and provides accurate life prediction results under low stress conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118965740A_ABST
    Figure CN118965740A_ABST
Patent Text Reader

Abstract

The invention provides a satellite-borne lithium ion battery life prediction method based on multi-working-condition degradation data. The method comprises the steps of 1, constructing a lithium ion battery degradation equivalent model based on a GBM process; 2, acquiring a likelihood function of the lithium ion battery degradation equivalent model; 3, based on the likelihood function of the lithium ion battery degradation equivalent model, combining the lithium ion battery degradation equivalent model to construct a satellite-borne lithium ion battery life prediction basic model; 4, performing linear migration on the satellite-borne lithium ion battery life prediction basic model to obtain a migration model; and step 5, obtaining a satellite-borne lithium ion battery life prediction result based on the migration model. The method can realize life prediction when the effective degradation sample of the battery is insufficient under the low-stress working condition. The multi-working-condition degradation equivalent model established by the linear migration method efficiently utilizes sufficient degradation historical data under other working conditions, and a theoretical basis is provided for predicting the service life of the satellite-borne lithium ion battery under the low-stress working condition within a short time.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a satellite-borne lithium-ion battery life prediction method based on multi-operating condition degradation data, and belongs to the technical field of lithium-ion batteries. Background Art

[0002] Batteries are an important part of satellite energy systems. Lithium-ion batteries have replaced nickel-cadmium batteries and nickel-hydrogen batteries and become the preferred battery type for satellite platforms due to their high energy density, long cycle life, safety and reliability. During the on-orbit operation of satellites, lithium-ion batteries store and release energy through a continuous charge and discharge process to ensure the normal operation of the satellite platform. However, lithium-ion batteries will undergo irreversible degradation during the charge and discharge cycle, resulting in continuous attenuation of indicators such as battery capacity and discharge power. Predicting the remaining life of onboard lithium-ion batteries during ground testing is an important step in ensuring the stable on-orbit operation and maintenance of satellite platforms.

[0003] By using the historical data of onboard lithium-ion battery degradation collected from accelerated degradation experiments in ground tests for equivalent modeling, we can infer the changing trend of battery health status with charge and discharge cycles, and then give a prediction result of the remaining battery life. However, under low stress conditions such as normal temperature, low charge and discharge rate, and low discharge depth, the battery will not degrade sufficiently in a short period of time, making it difficult to collect sufficient and effective degradation history data for life prediction.

[0004] The existing methods for predicting the remaining life of satellite-borne lithium-ion batteries under different working conditions have the following defects:

[0005] (1) The efficiency of battery degradation equivalent modeling is low, and the remaining life prediction cycle is long;

[0006] (2) Battery equivalent modeling relies on sufficiently effective historical degradation data from accelerated degradation experiments and cannot complete life prediction when effective samples are lacking under low stress conditions. Summary of the invention

[0007] The present invention aims to solve the technical problems in the prior art that the efficiency of equivalent modeling of lithium-ion battery degradation is low, the cycle required for remaining life prediction is long, and the lithium-ion battery life prediction cannot be completed when there is a lack of effective samples under low stress conditions. A satellite-borne lithium-ion battery life prediction method based on multi-condition degradation data is proposed.

[0008] The technical solution adopted by the present invention to solve the above problems is: the present invention proposes a method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data, comprising:

[0009] Step 1: Construct a lithium-ion battery degradation equivalent model based on the GBM process;

[0010] Step 2: Obtain the likelihood function of the equivalent model of lithium-ion battery degradation;

[0011] Step 3: Based on the likelihood function of the lithium-ion battery degradation equivalent model, the basic model for predicting the life of satellite-borne lithium-ion batteries is constructed in combination with the lithium-ion battery degradation equivalent model;

[0012] Step 4: Perform linear migration on the basic model for predicting the life of satellite-borne lithium-ion batteries to obtain a migration model;

[0013] Step 5: Obtain the prediction results of the onboard lithium-ion battery life based on the migration model.

[0014] Optionally, the steps of constructing the battery degradation equivalent model in step 1 include:

[0015] Step 1.1: Use the GBM process to model the degradation process of lithium-ion batteries and obtain a GBM process degradation model;

[0016] Step 1.2: Substitute the inverse power law model into the GBM process degradation model, and use other operating condition historical data sets to train the GBM process degradation model after substituting the inverse power law model to obtain a lithium-ion battery degradation equivalent model. The input of the lithium-ion battery degradation equivalent model is the SOH historical data of lithium-ion battery capacity degradation, and the output is the battery SOH prediction value;

[0017] The expression of GBM process degradation model is:

[0018]

[0019] In formula (1), C x is the remaining SOH of the battery when the battery has been charged and discharged for x number of cycles, C0 is the initial SOH of the battery, μ is the average growth rate of the battery SOH per unit time, σ is the fluctuation rate of the battery SOH, and W x is a Brownian motion with variance x;

[0020] The expression of the inverse power law model is:

[0021] μ=aS -b (2);

[0022] In formula (2), a and b are the parameter values ​​to be estimated, S is the single stress of the satellite-borne lithium-ion battery under different working conditions, and μ is the average growth rate;

[0023] The expression of the equivalent model of lithium-ion battery degradation is:

[0024]

[0025] In formula (3), σ, a, and b are all parameter values ​​to be estimated.

[0026] Optionally, the step of obtaining the likelihood function of the lithium-ion battery degradation equivalent model in step 2 includes:

[0027] Step 2.1: Substitute other operating condition historical data sets into the GBM process degradation model to obtain the likelihood function of the GBM process

[0028] Step 2.2: Obtain the likelihood function of the equivalent model of lithium-ion battery degradation based on the likelihood function of the GBM process;

[0029] The expression of the likelihood function of the equivalent model of lithium-ion battery degradation is:

[0030]

[0031] In formula (4), m is the total number of sample points used to calculate the likelihood function.

[0032] Optionally, the steps of constructing a basic model for predicting the life of a satellite-borne lithium-ion battery in step 3 include:

[0033] Based on the maximum likelihood estimation method, the likelihood function of the lithium-ion battery degradation equivalent model is identified, the maximum likelihood estimation values ​​of the parameters to be estimated σ, a, b of the lithium-ion battery degradation equivalent model are obtained, and the basic model for predicting the life of satellite-borne lithium-ion batteries is obtained;

[0034] The battery degradation modeling used in the present invention is simple, and the prediction accuracy of the life of satellite-borne lithium-ion batteries is high, while saving the time required for life prediction. The lithium-ion battery degradation equivalent model based on random processes has a simple structure, low training and computing resource overhead, and can complete the life prediction of a large number of battery models in a short time.

[0035] The maximum likelihood estimate of σ,a,b is calculated as:

[0036]

[0037] In formula (5), are the maximum likelihood estimates with estimated parameters σ, a, and b respectively.

[0038] Optionally, the steps for obtaining the migration model in step 4 include:

[0039] Based on the linear migration method and the historical data set of low-stress conditions, the basic model for predicting the life of onboard lithium-ion batteries was adjusted to obtain an initial migration model that meets the low-stress conditions.

[0040] The least square method is used to solve the linear transformation parameters to be estimated in the initial migration model to obtain the migration model;

[0041] The present invention applies a migration model, and the multi-condition degradation equivalent model established by the linear migration method effectively utilizes sufficient degradation history data under other conditions, and can predict the life of multiple types of lithium-ion batteries. When predicting the life of other types of lithium-ion batteries, it is only necessary to migrate the data of the corresponding type of lithium-ion battery to the basic model for life prediction of satellite-borne lithium-ion batteries, so as to realize the life prediction of the lithium-ion battery of this type, which provides a theoretical basis for the life prediction of satellite-borne lithium-ion batteries under low-stress conditions in a short time.

[0042] The expression of the migration model is:

[0043]

[0044] In formula (6), f(·) is the functional relationship between the charge-discharge cycle and the battery SOH in the model trained using other working conditions, and g(·) is the functional relationship obtained by linear migration that conforms to the equivalent model of battery degradation under low stress conditions. The remaining SOH of the battery at the number of charge and discharge cycles x under low stress conditions estimated by the migration model, a1 and b1 are the linear transformation parameters to be estimated;

[0045] The solution formula for the linear transformation parameters to be estimated is:

[0046]

[0047] In formula (7), C x ′ is the remaining SOH of the battery when the battery has x charge and discharge cycles under the actual low stress condition, and n is the total number of sample points of the historical data under the low stress condition.

[0048] Optionally, the step of obtaining the prediction result of the life of the satellite-borne lithium ions in step 5 includes:

[0049] Step 5.1: Based on the migration model, the number of cycles i in which the SOH is first predicted to be lower than the failure threshold for the first time in the charge and discharge cycle of the onboard lithium-ion battery is obtained, and the number of cycles in which the SOH is first predicted to be lower than the failure threshold for the first time is taken as the end point of the life of the onboard lithium-ion battery;

[0050] Step 5.2: Obtain the starting cycle number of the satellite-borne lithium-ion battery life prediction, and subtract the starting cycle number of the satellite-borne lithium-ion battery life prediction from the end point of the satellite-borne lithium-ion battery life to obtain the satellite-borne lithium-ion battery life prediction result;

[0051] The expression for the prediction result of the life of the onboard lithium-ion battery is:

[0052]

[0053] In formula (8), x RULis the predicted value of the remaining life of the onboard lithium-ion battery, x EOL The end of the life of the onboard lithium-ion battery. To predict the time when the SOH of the onboard lithium-ion battery falls below the failure threshold for the first time at the number of charge and discharge cycles i, t is the threshold coefficient, and x0 is the starting cycle number of the onboard lithium-ion battery life prediction.

[0054] The beneficial effects of the present invention are:

[0055] 1. The battery degradation modeling used in the present invention is simple, saving the time required for life prediction. The lithium-ion battery degradation equivalent model based on random processes has a simple structure, low training and computing resource overhead, and can complete the life prediction of a large number of battery models in a short time.

[0056] 2. The present invention can achieve life prediction when there are insufficient effective degradation samples of batteries under low stress conditions. The multi-condition degradation equivalent model established by the linear migration method efficiently utilizes sufficient degradation history data under other conditions, providing a theoretical basis for the life prediction of satellite-borne lithium-ion batteries under low stress conditions in a short period of time. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A flowchart of a method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data provided by the present invention;

[0058] Figure 2 A schematic diagram of a satellite-borne lithium-ion battery life prediction method based on multi-operating condition degradation data provided by the present invention;

[0059] Figure 3 A diagram showing the modeling results of the GBM process model using four sets of battery degradation sequence data in other operating condition historical data set 1 provided by the present invention;

[0060] Figure 4 A result diagram of model optimization using partial degradation sequences of two groups of battery cells in the low stress condition historical data set 2 provided by the present invention;

[0061] Figure 5 (a) and Figure 5 (b) are diagrams showing the prediction effects of the migration model provided by the present invention on the partial degradation sequences of two groups of battery cells in the low stress working condition historical data set 2. DETAILED DESCRIPTION

[0062] Combination Figure 1-2 This embodiment is described. In this embodiment, SOH is the health state of the battery, GBM is the geometric Brownian motion, and RUL is the remaining service life;

[0063] like Figure 1As shown, the method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data described in this embodiment includes:

[0064] S1: GBM process model initialization;

[0065] S101: Model the degradation process of lithium-ion batteries using the GBM process to obtain a GBM process degradation model;

[0066] S102: In order to describe the relationship between the single stress S of the battery under different working conditions and the average growth rate μ in the formula, an inverse power law model commonly used to describe the relationship between stress factors and product life is introduced into the GBM process degradation model. The GBM process degradation model after substituting the inverse power law model is trained using other working condition historical data sets 1 to obtain the lithium-ion battery degradation equivalent model 3;

[0067] The input of the lithium-ion battery degradation equivalent model is the SOH historical data of lithium-ion battery capacity degradation, and the output is the battery's SOH prediction value;

[0068] The expression of GBM process degradation model is:

[0069]

[0070] In formula (1), C x is the remaining SOH of the battery when the battery has been charged and discharged for x number of cycles, C0 is the initial SOH of the battery, μ is the average growth rate of the battery SOH per unit time, σ is the fluctuation rate of the battery SOH, and W x is a Brownian motion with variance x;

[0071] The expression of the inverse power law model is:

[0072] μ=aS -b (2);

[0073] In formula (2), a and b are parameter values ​​to be estimated;

[0074] The expression of the equivalent model of lithium-ion battery degradation is:

[0075]

[0076] In formula (3), σ, a, and b are all parameter values ​​to be estimated.

[0077] like Figure 3 As shown, this embodiment uses four groups of battery degradation sequence data in other operating condition historical data set 1 to model the GBM process model.

[0078] S2: Model likelihood function representation;

[0079] S201: Figure 2 As shown, the likelihood function of the GBM process is obtained by substituting other operating condition historical data sets 1 into the GBM process degradation model;

[0080] S202: Obtaining a likelihood function of a lithium-ion battery degradation equivalent model based on a likelihood function of a GBM process;

[0081] The expression of the likelihood function of the equivalent model of lithium-ion battery degradation is:

[0082]

[0083] In formula (4), m is the total number of sample points used to calculate the likelihood function.

[0084] S3: Estimation of unknown parameters of the basic model for predicting the life of satellite-borne lithium-ion batteries;

[0085] Based on the maximum likelihood estimation method, the likelihood function of the lithium-ion battery degradation equivalent model is identified, the maximum likelihood estimation values ​​of the parameters to be estimated σ, a, b of the lithium-ion battery degradation equivalent model are obtained, and the basic model 4 for predicting the life of satellite-borne lithium-ion batteries is obtained;

[0086] The maximum likelihood estimate of σ,a,b is calculated as:

[0087]

[0088] In formula (5), are the maximum likelihood estimates with estimated parameters σ, a, and b respectively.

[0089] In this embodiment, after the likelihood function is optimized, the values ​​of the unknown parameters in the basic model for predicting the life of satellite-borne lithium-ion batteries are determined to be σ: 6.125e-04, a: 2.599e-04, and b: -4.742e-01.

[0090] S4: Linear migration of basic models for predicting life of spaceborne lithium-ion batteries;

[0091] Based on the high-precision degradation model trained by the historical dataset of satellite-borne lithium-ion batteries under other working conditions1, a multi-condition degradation equivalent model is established by adjusting the output through the linear migration method, so that the migration model can adapt to the battery degradation characteristics under low stress conditions.

[0092] S401: adjusting the basic model 4 for predicting the life of a satellite-borne lithium-ion battery based on a linear migration method and a historical data set 2 of low stress conditions to obtain an initial migration model that meets the low stress conditions;

[0093] S402: using the least square method to solve the linear transformation parameters to be estimated in the initial migration model to obtain a migration model 5;

[0094] The expression of the migration model is:

[0095]

[0096] In formula (6), f(·) is the functional relationship between the charge-discharge cycle and the battery SOH in the model trained using other working conditions, and g(·) is the functional relationship obtained by linear migration that conforms to the equivalent model of battery degradation under low stress conditions. The remaining SOH of the battery at the number of charge and discharge cycles x under low stress conditions estimated by the migration model, a1 and b1 are the linear transformation parameters to be estimated;

[0097] The solution formula for the linear transformation parameters to be estimated is:

[0098]

[0099] In formula (7), C x ′ is the remaining SOH of the battery when the battery has x charge and discharge cycles under the actual low stress condition, and n is the total number of sample points of the historical data under the low stress condition.

[0100] like Figure 2 As shown, in this embodiment, partial degradation sequences of two groups of battery cells in the low stress condition historical data set 2 are used to tune the model, and the estimated values ​​of the unknown parameters using the least squares method are: a1: 4.844e-01, b1: 4.689e-01.

[0101] S5: Prediction of remaining life of onboard lithium-ion batteries;

[0102] Based on the migration model, the predicted value of the SOH of the satellite-borne lithium-ion battery under low stress conditions with the cycle change of the charge and discharge points is obtained. The cycle number at which the predicted SOH of the satellite-borne lithium-ion battery first drops to a certain proportion of the initial value is taken as the end of the life of the satellite-borne lithium-ion battery. The end of the life of the satellite-borne lithium-ion battery is subtracted from the predicted starting cycle number to obtain the RUL prediction result of the satellite-borne lithium-ion battery 6;

[0103] The expression for the prediction result of the life of the onboard lithium-ion battery is:

[0104]

[0105] In formula (8), x RUL is the predicted value of the remaining life of the onboard lithium-ion battery, x EOL The end of the life of the onboard lithium-ion battery. To predict the time when the SOH of the onboard lithium-ion battery falls below the failure threshold for the first time at the number of charge and discharge cycles i, t is the threshold coefficient, and x0 is the starting cycle number of the onboard lithium-ion battery life prediction.

[0106] like Figure 5 As shown in (a) and (b), the average RUL prediction error of the two groups of battery cells in the low stress condition historical data set 2 is 3.68%. In this embodiment, based on the programming of Python language, the unknown parameter estimation of the basic model for lithium-ion battery life prediction takes about 2 minutes, and the linear migration of the model takes about 1 minute. For a set of battery degradation sequences under a specific condition, it only takes about 3 minutes to achieve its high-precision RUL prediction task, and the life prediction of a large number of battery models can be completed in a short time.

[0107] The above description is only a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any technician familiar with the profession can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent replacement and improvement of the above embodiments made according to the technical essence of the present invention, within the spirit and principles of the present invention, still fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data, characterized in that: The method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data comprises the following steps: Step 1: Construct a lithium-ion battery degradation equivalent model based on the GBM process; Step 2: Obtain the likelihood function of the equivalent model of lithium-ion battery degradation; Step 3: Based on the likelihood function of the lithium-ion battery degradation equivalent model, a basic model for predicting the life of a satellite-borne lithium-ion battery is constructed in combination with the lithium-ion battery degradation equivalent model; Step 4: linearly migrate the basic model for predicting the life of the satellite-borne lithium-ion battery to obtain a migration model; Step 5: Obtain the life prediction result of the onboard lithium-ion battery based on the migration model.

2. The method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data according to claim 1, characterized in that: The steps of constructing the battery degradation equivalent model in step 1 include: Step 1.1: Use the GBM process to model the degradation process of lithium-ion batteries and obtain a GBM process degradation model; Step 1.2: Substitute the inverse power law model into the GBM process degradation model, and use other operating condition historical data sets to train the GBM process degradation model after substituting the inverse power law model to obtain a lithium-ion battery degradation equivalent model, wherein the input of the lithium-ion battery degradation equivalent model is the SOH historical data of lithium-ion battery capacity degradation, and the output is the battery SOH prediction value; The expression of GBM process degradation model is: In formula (1), C x is the remaining SOH of the battery when the battery has been charged and discharged for x number of cycles, C0 is the initial SOH of the battery, μ is the average growth rate of the battery SOH per unit time, σ is the fluctuation rate of the battery SOH, and W x is a Brownian motion with variance x; The expression of the inverse power law model is: μ=aS -b (2); In formula (2), a and b are parameter values ​​to be estimated; The expression of the equivalent model of lithium-ion battery degradation is: In formula (3), σ, a, and b are all parameter values ​​to be estimated.

3. The method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data according to claim 1, characterized in that: The steps of obtaining the likelihood function of the lithium-ion battery degradation equivalent model in step 2 include: Step 2.1: Substitute other operating condition historical data sets into the GBM process degradation model to obtain the likelihood function of the GBM process Step 2.2: Obtain the likelihood function of the equivalent model of lithium-ion battery degradation based on the likelihood function of the GBM process; The expression of the likelihood function of the equivalent model of lithium-ion battery degradation is: In formula (4), m is the total number of sample points used to calculate the likelihood function.

4. The method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data according to claim 1, characterized in that: The steps of constructing the basic model for predicting the life of onboard lithium-ion batteries in step 3 include: Based on the maximum likelihood estimation method, the likelihood function of the lithium-ion battery degradation equivalent model is identified, the maximum likelihood estimation values ​​of the parameters to be estimated σ, a, b of the lithium-ion battery degradation equivalent model are obtained, and the basic model for predicting the life of satellite-borne lithium-ion batteries is obtained; The maximum likelihood estimate of σ,a,b is calculated as: In formula (5), are the maximum likelihood estimates with estimated parameters σ, a, and b respectively.

5. The method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data according to claim 1, characterized in that: The steps to obtain the migration model in step 4 include: Step 4.1: Based on the linear migration method and the historical data set of low stress conditions, the basic model for predicting the life of the onboard lithium-ion battery is adjusted to obtain an initial migration model that meets the low stress conditions; Step 4.2: Use the least square method to solve the linear transformation parameters to be estimated in the initial migration model to obtain the migration model; The expression of the migration model is: In formula (6), f(·) is the functional relationship between the charge-discharge cycle and the battery SOH in the model trained using other working conditions, and g(·) is the functional relationship obtained by linear migration that conforms to the equivalent model of battery degradation under low stress conditions. The remaining SOH of the battery at the number of charge and discharge cycles x under low stress conditions estimated by the migration model, a1 and b1 are the linear transformation parameters to be estimated; The solution formula for the linear transformation parameters to be estimated is: In formula (7), C x ′ is the remaining SOH of the battery when the battery has x charge and discharge cycles under the actual low stress condition, and n is the total number of sample points of the historical data under the low stress condition.

6. The method for predicting the life of a satellite-borne lithium-ion battery based on multi-operating condition degradation data according to claim 1, characterized in that: The steps of obtaining the prediction results of the onboard lithium ion life in step 5 include: Step 5.1: Based on the migration model, the number of cycles i in which the SOH is first predicted to be lower than the failure threshold for the first time in the charge and discharge cycle of the onboard lithium-ion battery is obtained, and the number of cycles in which the SOH is first predicted to be lower than the failure threshold for the first time is taken as the end point of the life of the onboard lithium-ion battery; Step 5.2: Obtain the starting cycle number of the satellite lithium-ion battery life prediction, and subtract the starting cycle number of the satellite lithium-ion battery life prediction from the end point of the satellite lithium-ion battery life to obtain the satellite lithium-ion battery life prediction result; The expression for the prediction result of the life of the onboard lithium-ion battery is: In formula (8), x RUL is the predicted value of the remaining life of the onboard lithium-ion battery, x EOL The end of the life of the onboard lithium-ion battery. To predict the time when the SOH of the onboard lithium-ion battery falls below the failure threshold for the first time at the number of charge and discharge cycles i, t is the threshold coefficient, and x0 is the starting cycle number of the onboard lithium-ion battery life prediction.