A battery life model parameter fitting method

Abstract

The present invention discloses a battery life model parameter fitting method, comprising the following steps: S1: statistically analyzing the battery storage time and battery cumulative charge throughput data of a durability test vehicle; S2: calculating the battery storage time data and the fitting formula Qfade according to the battery storage time data and the fitting formula Qfade Cale , use the heuristic algorithm to fit the parameters A1, A2, and A3 to be identified; S3: according to the battery cumulative charge throughput data and the fitting formula Qfade cyc , and the heuristic algorithm is used to fit the parameters to be identified B1, B2, and B3. The method described in the present invention directly uses the vehicle durability test data to fit the life model parameters, which can well solve the error caused by the traditional method of using fixed constant temperature conditions to fit the parameters of the vehicle working condition. At the same time, the vehicle durability test data can be directly used to fit the battery life model parameters, saving millions of battery life test costs.

Description

Technical Field

[0001] The present invention relates to the field of battery life prediction, and in particular to a battery life model parameter fitting method. Background Art

[0002] The traditional battery life model mainly fits the battery life model parameters based on the cycle and calendar life test data at different constant temperatures. However, this method cannot truly be equivalent to the various working conditions during the use of the entire vehicle. Therefore, the fitting effect of the traditional method is generally very good, but the effect is very general when applied to the entire vehicle.

[0003] For example, a "life prediction method for a lithium-ion battery aging model" disclosed in a Chinese patent document, with the publication number CN110320474A, includes the following steps: (1) testing the capacity retention rate of lithium batteries in a 50% SOC state after storage at different temperatures and the value of the battery DC internal resistance changing with storage time; (2) storing the batteries in different voltage states at the same temperature for the same time, and performing AC impedance test and DC internal resistance test; (3) analyzing the correlation between calendar life capacity decay, AC internal resistance and DC internal resistance change and temperature and voltage, and describing the change process of experimental aging data over time through fitting functions; (4) analyzing the model based on the fitted data, fitting the actual measured effective data, and effectively predicting the future battery calendar life and battery internal resistance parameters. Although the test method of this invention is relatively simple, does not require high-end equipment and complex calculations, and greatly shortens the test cycle, this method is mainly based on fitting the battery life model parameters based on the cycle and calendar life test data at different constant temperatures, and does not solve the problem of not being able to truly equate the various different working conditions during the use of the whole vehicle. Summary of the invention

[0004] The present invention aims to overcome the problem that the prior art cannot truly equate the various working conditions during the use of the whole vehicle, so the fitting effect is very good, but the effect is very general when applied to the whole vehicle, and provides a battery life model parameter fitting method.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A battery life model parameter fitting method, comprising the following steps: S1: statistically analyzing the battery storage time and battery cumulative charge throughput data of a durability test vehicle; S2: calculating the battery storage time data and the fitting formula Qfade according to the battery storage time data and the fitting formula Qfade Cale , use the heuristic algorithm to identify the parameter A 1 , A 2 , A 3 S3: Based on the battery cumulative charge throughput data and the fitting formula Qfadecyc , use the heuristic algorithm to identify the parameter B 1 , B 2 , B 3 The method of the present invention directly uses the vehicle durability test data to fit the life model parameters, which can well solve the error caused by the traditional method of using fixed constant temperature working conditions to fit the parameters of the vehicle working conditions. At the same time, the vehicle durability test data can be directly used to fit the battery life model parameters, saving millions of battery life test costs.

[0007] As a preferred solution of the present invention, the battery storage time data in S1 is the battery storage time data of the durability test vehicle at different temperatures. The present invention uses the vehicle durability test data to fit the life model parameters, which can well solve the error caused by the traditional method of using the fixed constant temperature working condition to fit the parameters of the vehicle working condition, and solves the problem that the fitting effect is very good, but the effect is very general when applied to the whole vehicle.

[0008] As a preferred solution of the present invention, the battery cumulative charge throughput data in S1 is the battery cumulative charge throughput data of the endurance test vehicle at different temperatures. The present invention uses the vehicle endurance test data to fit the life model parameters, which can well solve the error caused by the traditional method of using the fixed constant temperature working condition to fit the parameters of the vehicle working condition, and solves the problem that the fitting effect is very good, but the effect is very general when applied to the whole vehicle.

[0009] As a preferred embodiment of the present invention, the fitting formula Qfade in S2 is Cale Specifically:

[0010]

[0011] Among them, A 1 , A 2 , A 3 is the parameter to be identified, T 1 , T 2 …T n To test the vehicle's durability at different ambient temperatures, t 1 ,t 2 …t n is the battery storage time corresponding to the durability test vehicle at different ambient temperatures, and t is the sum of the battery storage time corresponding to the durability test vehicle at different temperatures, that is, t = t 1 +t 2 +…+t n The existing methods are more inclined to academic research, without considering the complexity of the engineering application environment. The specific constant temperature experimental environment is equivalent to the complex temperature environment of the whole vehicle, and the effect is relatively poor. The present invention directly uses the whole vehicle data to fit the battery life, which has higher accuracy.

[0012] As a preferred embodiment of the present invention, the fitting formula Qfade in S3 is cyc Specifically:

[0013]

[0014] Among them, B 1 , B 2 , B 3 is the parameter to be identified, T 1 , T 2 …T n To test the durability of the vehicle in different ambient temperatures, Ah 1 Ah 2 …Ah n is the battery cumulative charge throughput corresponding to the endurance test vehicle at different temperatures, Ah is the sum of the battery cumulative charge throughput corresponding to the endurance test vehicle at different ambient temperatures, that is, Ah = Ah 1 +Ah 2 +…+Ah n The existing methods are more inclined to academic research, without considering the complexity of the engineering application environment. The specific constant temperature experimental environment is equivalent to the complex temperature environment of the whole vehicle, and the effect is relatively poor. The present invention directly uses the whole vehicle data to fit the battery life, which has higher accuracy.

[0015] As a preferred solution of the present invention, the heuristic algorithm in S2 and S3 is specifically the PS0 algorithm. The PS0 algorithm is a kind of heuristic algorithm, namely the particle swarm optimization algorithm. The heuristic algorithm also includes simulated annealing algorithm (SA), genetic algorithm (GA), ant colony algorithm (ACO), artificial neural network (ANN), etc.

[0016] Therefore, the present invention has the following beneficial effects: the method described in the present invention directly uses the durability test data of the whole vehicle to fit the life model parameters, which can well solve the errors caused by the traditional method of using the fixed constant temperature working condition to fit the parameters of the whole vehicle working condition equivalent, and at the same time, the battery life model parameters can be directly fitted with the whole vehicle durability test data, saving millions of battery life test costs; the present invention uses the durability test data of the whole vehicle to fit the life model parameters, which can well solve the errors caused by the traditional method of using the fixed constant temperature working condition to fit the parameters of the whole vehicle working condition equivalent, and solves the problem that the fitting effect is very good, but the effect is very general when applied to the whole vehicle; the present invention directly uses the whole vehicle data to fit the battery life, with higher accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of the method of the present invention;

[0018] Figure 2 It is a graph of the cumulative Ah amount and the corresponding equivalent cycle number at different temperatures according to the statistics of the present invention;

[0019] Figure 3 It is a fitting effect diagram of battery capacity attenuation parameters caused by battery storage time of the present invention. DETAILED DESCRIPTION

[0020] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0021] Battery capacity attenuation is mainly composed of two parts: storage time and cycle. Among them, the commonly used fitting formula for battery capacity attenuation caused by common storage time is as follows:

[0022]

[0023] Among them, A 1 , A 2 , A 3 is the parameter to be identified, T is the ambient temperature, and t is the storage time.

[0024] The commonly used fitting formula for battery capacity attenuation caused by existing common cycles is as follows:

[0025]

[0026] Among them, B 1 , B 2 , B 3 are the parameters to be identified, T represents the ambient temperature, and Ah represents the accumulated charge throughput.

[0027] Battery capacity health status: SOH = 1-Qfade Cale -Qfade cyc .

[0028] The existing methods are more inclined to academic research, without considering the complexity of the engineering application environment. Using a specific constant temperature experimental environment to equal the complex temperature environment of the whole vehicle has a relatively poor effect. 1 , A 2 , A 3 , B 1 , B 2 , B 3 All of them use the single variable method to test the influence of different temperatures on the health status of battery capacity. This method leads to the problem that the data used for fitting parameters is quite different from the whole vehicle data, resulting in model errors. To solve this problem, a battery life model parameter fitting method is proposed. The present invention directly uses the whole vehicle data to fit the battery life, which has higher accuracy. Figure 1 As shown, the following steps are included: S1: Counting the battery storage time and battery cumulative charge throughput data of the endurance test vehicle; S2: Based on the battery storage time data and the fitting formula Qfade Cale , use the heuristic algorithm to identify the parameter A1 , A 2 , A 3 S3: Based on the battery cumulative charge throughput data and the fitting formula Qfade cyc , use the heuristic algorithm to identify the parameter B 1 , B 2 , B 3 The fitting of

[0029] This embodiment takes the capacity attenuation formula caused by cycling as an example: First, the cumulative Ah amount of the endurance test vehicle at different temperatures is counted, as shown in the following table:

[0030] T(℃) <![CDATA[T 1 ]]> <![CDATA[T 2 ]]> … <![CDATA[T n ]]> Ah <![CDATA[Ah 1 ]]> <![CDATA[Ah 2 ]]> … <![CDATA[Ah n ]]>

[0031] Use the following formula to get the equivalent Qfade cyc Battery capacity attenuation formula:

[0032]

[0033] Among them, B 1 , B 2 , B 3 is the parameter to be identified, T 1 , T 2 …T n The different ambient temperatures of the endurance test vehicle, in degrees Celsius, Ah 1 Ah 2 …Ah n is the battery cumulative charge throughput corresponding to the endurance test vehicle at different temperatures, Ah is the sum of the battery cumulative charge throughput corresponding to the endurance test vehicle at different ambient temperatures, that is, Ah = Ah 1 +Ah 2 +…+Ah n .

[0034] According to the equivalent formula and the statistical vehicle durability data, the PSO algorithm is used in this embodiment to fit the parameter B to be identified by using the heuristic algorithm. 1 , B 2 , B 3 .

[0035] Based on the battery pack simulated vehicle operating life test data, the cumulative Ah and the corresponding equivalent cycle number at different temperatures are statistically shown in the following table and Figure 2 As shown:

[0036] T(℃) 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 Ah 4266.09 17037.6 53355.8 45445.4 400961 3211758 1.2E+07 1.3E+07 1.4E+07 2271913 2395113 6151958 6804632 7300577 7181285 3410476 1158682

[0037] According to the above data and Qfade cvc Formula, using the PSO algorithm in the heuristic algorithm to fit B 1 =948, B2 =5008,B 3 =0.6347, the fitting effect is as follows Figure 3 shown.

[0038] The fitting method of the battery capacity attenuation parameters caused by storage time is the same as the above method and will not be described in detail.

[0039] This embodiment only provides a method for equivalently fitting battery life parameters, using a relatively common life model formula. By using other similar life formulas and the equivalent method mentioned in the present invention, it is also possible to fit the battery life model parameters according to the vehicle operation data.

[0040] The method described in the present invention uses the whole vehicle durability test data to fit the battery life model parameters, reduces the workload of life testing, and saves millions of battery life testing costs. It can well solve the errors caused by the traditional method of fitting parameters equivalent to the whole vehicle working condition with a fixed constant temperature working condition. At the same time, the battery life model parameters can be directly fitted with the whole vehicle durability test data, and the actual operating parameters of the cloud customers can be used in the later stage to optimize the battery life model parameters.

[0041] The present invention uses the vehicle durability test data to fit the life model parameters, which can effectively solve the error caused by the traditional method of using fixed constant temperature working conditions to fit the parameters of the vehicle working conditions, and solves the problem that the fitting effect is very good, but the effect is very general when applied to the whole vehicle.

[0042] The present invention fits the battery life model parameters according to the actual working conditions of the whole vehicle, which is more in line with the actual usage and has higher accuracy.

[0043] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto, and any changes or substitutions that are not conceived through creative work should be included in the protection scope of the present invention.

Claims

1. A battery life model parameter fitting method, characterized in that: The following steps are involved: S1: Statistics of battery storage time and battery cumulative charge throughput of endurance test vehicles; S2: Based on the battery storage time data and fitting formula Qfade Cale , use the heuristic algorithm to fit the parameters A1, A2, and A3 to be identified; S3: Based on the battery cumulative charge throughput data and the fitting formula Qfade cyc , use the heuristic algorithm to fit the parameters B1, B2, and B3 to be identified; Qfade Cale for: A1, A2, A3 are parameters to be identified, T1, T2…T n For the different ambient temperatures of the endurance test vehicle, t1, t2…t n is the battery storage time corresponding to the endurance test vehicle at different ambient temperatures, and t is the sum of the battery storage time corresponding to the endurance test vehicle at different temperatures.

2. A battery life model parameter fitting method according to claim 1, characterized in that: The battery storage time data in S1 is the battery storage time data of the endurance test vehicle at different temperatures.

3. The battery life model parameter fitting method according to claim 1, characterized in that: The battery cumulative charge throughput data in S1 is the battery cumulative charge throughput data of the endurance test vehicle at different temperatures.

4. A battery life model parameter fitting method according to claim 1 or 3, characterized in that: The fitting formula Qfade in S3 cyc Specifically: Among them, B1, B2, B3 are parameters to be identified, T1, T2…T n To test the durability of the vehicle in different ambient temperatures, Ah1, Ah2…Ah n is the battery cumulative charge throughput corresponding to the endurance test vehicle at different temperatures, Ah is the sum of the battery cumulative charge throughput corresponding to the endurance test vehicle at different ambient temperatures, that is, Ah = Ah1 + Ah2 + ... + Ah n .

5. The battery life model parameter fitting method according to claim 1, characterized in that: The heuristic algorithm in S2 and S3 is specifically the PS0 algorithm.

Images