Method for predicting and evaluating service life of long-service-life lithium ion battery

Abstract

The invention belongs to the technical field of lithium ion batteries, and particularly relates to a method for predicting and evaluating the service life of a long-service-life lithium ion battery. Taking out 3n lithium ion batteries manufactured under the same condition, and testing initial capacity, internal resistance, voltage and electrochemical impedance spectroscopy (EIS) at 25 DEG C; circulating the lithium ion battery under the temperature condition of T1 / T2 / T3 / ... / Tn, and testing the EIS impedance, capacity and internal resistance when the lithium ion battery is circulated for 100 times, 500 times, 1000 times and 2000 times to 80% of the initial capacity of the battery; 3n is only placed under the temperature condition of T1 / T2 / T3 / ... / Tn, n is larger than 3, the placing time is 0 h, 500 h, 1000 h, 1500 h and 2000 h, and the direct current internal resistance (DCR), the capacity and the internal resistance (ACR) of the battery are measured in the placing process; and adding a logic formula for fitting to perform data learning calculation, confirming the capacity attenuation degree and impedance increase according to the use condition requirement to confirm a service life termination node, and defaulting that the service life of the battery is terminated when the capacity is required to be attenuated to 70% of the nominal capacity. Through the method, the lithium ion battery life prediction method is obtained, and the method has the advantages of low cost, simple equipment requirement and short preparation process, and has the capability of correctly evaluating the life of the long-life lithium ion battery.

Description

Technical Field

[0001] The invention belongs to the technical field of lithium ion batteries, and in particular relates to a life prediction and evaluation method for a long-life lithium ion battery. Background Art

[0002] Lithium-ion batteries are widely used as the main energy storage devices in military electronic products, avionics, electric vehicles, and various portable electronic devices (such as laptops, digital cameras, tablet computers, mobile phones, etc.) due to their long life, fast charging, high energy, small size, and no pollution. In actual application, the capacity of lithium-ion batteries will decrease with the increase of charge and discharge cycles, and the performance will gradually deteriorate. Battery life failure may lead to safety problems, so battery life prediction is particularly important.

[0003] The life degradation of lithium-ion batteries objectively exists throughout their entire life cycle. Its life problem mainly refers to the gradual degradation of the physicochemical structural properties of the positive and negative active materials that affect its discharge capacity, the bonding strength of the binder to the coating, the quality of the diaphragm, etc. during the cyclic charge and discharge process. Unexpected end of battery life often leads to failure of the entire system. Therefore, degradation prediction and analysis of batteries can provide timely and effective maintenance measures and battery replacement decisions, which is of great significance to improving system reliability and preventing catastrophic accidents. During the long-term use of lithium-ion batteries, the discharge capacity of lithium iron phosphate batteries gradually decreases due to a series of physical and chemical changes that occur inside them, that is, the health state (State of health, state of charge) of the battery gradually decreases. The internal resistance of the battery measured by impedance spectroscopy changes little during the battery degradation process, and its impact on the entire system is small and easy to be ignored, but it shows the degree of irreversible reaction of the internal structure of the battery from a microscopic perspective.

[0004] The prediction methods of the remaining life of lithium-ion batteries can be divided into three categories: model-based methods, data-driven methods, and fusion methods. Model-based methods can better reflect the physical and electrochemical characteristics of batteries, but it is difficult for model-based methods to monitor the internal state of batteries and accurate physical models are usually difficult to obtain. Model-based methods usually use prior knowledge of the product life cycle to construct mathematical functions to describe the physical characteristics and failure modes of the system. On this basis, a mathematical model that can reflect the physical laws of system performance degradation is established to deeply understand the essence of the object system and obtain more accurate prediction results. Commonly used methods include Kalman filtering, extended Kalman filtering, and particle filtering. Data-driven methods are a more popular method because of their flexibility and ease of operation. However, data-driven methods rely heavily on data. For example, uncertainty or incompleteness of data will significantly affect the performance of data-driven methods. Data-driven methods usually extract typical data from sensor data (e.g., voltage, current, temperature, time).

[0005] It is usually difficult for a single method to accurately describe the nonlinearity of battery degradation and fully adapt to the changing battery operating conditions. Moreover, the prediction of lithium-ion battery life abroad mainly focuses on modeling and algorithms, without comprehensive evaluation based on actual operating conditions. The accuracy needs to be improved. It is necessary to combine manual experience and comprehensive algorithm evaluation to solve the life assessment of long-life lithium-ion batteries, especially the evaluation of ultra-long-life batteries. Summary of the invention

[0006] In view of the above problems, the present invention provides a long-life lithium-ion battery life prediction and evaluation method, which is a lithium-ion battery life prediction method based on measured data such as battery impedance, high temperature aging and calendar storage, and data learning, and is compared in parallel with the measured values ​​to improve the prediction accuracy.

[0007] The long-life lithium-ion battery life prediction and evaluation method of the present invention comprises the following specific steps: (1) Take out 3n lithium-ion batteries manufactured under the same conditions and test the initial capacity, internal resistance, voltage and electrochemical impedance spectroscopy (EIS) at 25°C; (2) The lithium-ion battery is cycled under the temperature conditions of T1 / T2 / T3 / ··· / Tn, where n is greater than 3 and the cycle current is the nominal current of the battery. The test cycles are 100, 500, 1000 and 2000 times, and the EIS impedance, capacity and internal resistance are measured when the battery reaches 80% of its initial capacity; (3) Take 3n batteries of the same model as (1) and place them under the temperature conditions of T1 / T2 / T3 / ·· / Tn, where n is greater than 3, and the placing time is 0h, 500h, 1000h, 1500h and 2000h. During this period, measure the DC internal resistance (DCR), capacity and internal resistance (ACR) of the battery; (4) Add the logical formula for fitting to the data in steps (2) and (3), merge and fit according to the usage conditions of different batteries. According to the type and design of the battery, the values ​​of R0 to R4 in the EIS four-circuit method that change with degradation are most different. The four-circuit method is used to express R0: battery internal resistance (SEI resistance), R1: positive electrode internal resistance; R2: negative electrode internal resistance; R3: electrolyte diffusion resistance; R0 / R1 / R2 / R3, Warburg coefficient (σ) represents the slope of the fitting line, ω represents the angular frequency in the low frequency region, Li + The diffusion coefficient is closely related to the Warburg coefficient (σ)

[191] , the fitting calculation formula is as follows: Z' = Re + Rct + σω −0.5 (1) D Li+ = R 2 T 2 / 2A 2 n 4 F 4 C 2 σ 2 (2) Where R is the ideal gas constant, T is the absolute temperature, A is the active area of ​​the electrode, n is the number of electrons lost or reduced in each molecule, F is the Faraday constant, C is the molar concentration of lithium ions intercalated in the LFP particles, and σ is the Warburg coefficient. -0.5 The linear fitting of is obtained.

[0008] Before performing data learning calculations (Formula 1 and Formula 2), take batteries of the same model and repeat the test according to steps (1) to (4); The capacity attenuation degree and impedance increase are confirmed according to the operating conditions to confirm the end-of-life node. The battery life is terminated by default when the capacity decays to 70% of the nominal capacity. According to the method for predicting and evaluating the life of a long-life lithium-ion battery described in the above, the batteries with the same manufacturing conditions described in step (1) are products of the same structure manufactured using the same or similar equipment, and the positive electrode material, negative electrode material, electrolyte, foil and corresponding auxiliary materials used are the same.

[0009] According to the long-life lithium-ion battery life prediction and evaluation method described in the above, in step (2), T1 / T2 / T3 / ··· / Tn are selected based on the battery design temperature usage range including the upper limit, lower limit and usage conditions of the battery design temperature, and are the objects for segmented test evaluation.

[0010] According to the long-life lithium-ion battery life prediction and evaluation method described above, the current in step (2) is selected according to the designed battery nominal current cycle and is adjusted according to the current actually used by the battery.

[0011] According to the long-life lithium-ion battery life prediction and evaluation method described in, in step (3), the test formula of the direct current internal resistance (DCR) is DCR=(U1-U2) / (I2-I1), U1 is the end voltage of step (2), U2 is the end voltage of step (3), I1 is the discharge current of step (2), and I2 is the discharge current of step (3).

[0012] According to the long-life lithium-ion battery life prediction and evaluation method described in the above, the battery capacity and impedance data are calculated using the fitting logic described in step (4) to fit the attenuation logic formula.

[0013] According to the long-life lithium-ion battery life prediction and evaluation method described in the above, the repetitive operation of step (5) is to modify and learn the impedance data calculated by the logic of step (4), and summarize the change trend of the battery system and fit the formula based on the growth law of the four resistances R0, R1, R2, and R3, the capacity decay rate, the charge and discharge time, the battery static time and other parameters.

[0014] Compared with the existing accelerated aging methods, the essential characteristics and creativity of the present invention are embodied in: 1. Through the method of the present invention, a method for predicting the life of a lithium-ion battery is obtained. The method has the advantages of low cost, simple equipment requirements, and short preparation steps, and has the ability to correctly evaluate the life of a long-life lithium-ion battery.

[0015] 2. The present invention accelerates the actual life evaluation of long-life lithium-ion batteries by creatively using different attenuation rates of batteries under different temperature conditions, provides a practical application evaluation method, and provides data support for the maintenance cycle in product terminal applications. Temperature selection, including the upper and lower temperature limits of theoretical design, is set according to the size of the electrochemical window of the battery design through a combination of theoretical design and actual measurement.

[0016] 3. In the present invention, continuous charge and discharge cycle and static aging are used for evaluation. The actual battery usage conditions can be weighted and calculated, including modes of different usage conditions, with high compatibility.

[0017] 4. The present invention uses repeated testing to modify and learn the current logic. The accuracy gradually converges and improves. In addition, the measured data can be predicted with higher accuracy, which is different from pure logical operations or data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 EIS curve of 18650 lithium iron phosphate battery in Example 1 Figure 2 Life fitting curve of 18650 lithium iron phosphate battery DETAILED DESCRIPTION

[0019] The present invention is further described below in conjunction with specific examples, but the present invention is not limited to the following examples, and the methods described are all conventional methods unless otherwise specified.

[0020] The specific implementation process of the long-life lithium-ion battery life prediction and evaluation method of the present invention includes the following steps: First, 3 lithium-ion batteries manufactured under the same conditions were tested at 25°C for initial capacity, internal resistance, voltage and electrochemical impedance spectroscopy (EIS); batteries manufactured under the same manufacturing conditions are products of the same structure manufactured using the same or similar equipment, and there are no major material differences in the positive electrode materials, negative electrode materials, electrolytes, foils and corresponding auxiliary materials.

[0021] The lithium-ion battery is cycled under three or more temperature conditions of T1 / T2 / T3 / Tn, and the cycle current is the nominal current of the battery; the test cycle is 100 times, 500 times, and 1000 times, and the cycle is cycled to 80% of the initial capacity of the battery; then the EIS, capacity, and internal resistance of the battery are obtained at 1 / T2 / T3 / Tn and when the battery is cycled to 80% of the initial capacity (capacity retention rate SOH is 80%).

[0022] These are the objects of segmented test evaluation based on the size of the electrochemical window of the battery design, including the upper and lower limits and usage conditions. The current is selected according to the designed battery nominal current cycle and adjusted according to the current actually used by the battery.

[0023] Take 3n batteries from the same batch and place them under T1 / T2 / T3 / Tn temperature conditions for 0h, 500h, 1000h, 1500h, 2000h; during this period, the DC internal resistance (DCR), capacity and internal resistance (ACR) of the batteries are tested; the above test data are added to the logic for fitting respectively, and combined for fitting according to different battery usage conditions.

[0024] Process data diagram of the above steps The logical formulas for fitting are added to the above data respectively, and they are merged and fitted according to the usage conditions of different batteries. According to the type and design of the battery, the values ​​of R0 ~ R4 in the EIS four-circuit method, which change with degradation, are most different. The four-circuit method is used to express R0: battery internal resistance (SEI resistance), R1: positive electrode internal resistance; R2: negative electrode internal resistance; R3: electrolyte diffusion resistance; R0 / R1 / R2 / R3, Warburg coefficient (σ) represents the slope of the fitting line, ω represents the angular frequency in the low frequency region, Li + The diffusion coefficient is closely related to the Warburg coefficient (σ)

[191] , the fitting calculation formula is as follows: Z' = Re + Rct + σω −0.5 (1) D Li + = R 2 T 2 / 2A 2 n4 F 4 C 2 σ 2 (2) Where R is the ideal gas constant, T is the absolute temperature, A is the active area of ​​the electrode, n is the number of electrons lost or reduced in each molecule, F is the Faraday constant, C is the molar concentration of lithium ions intercalated in the LFP particles, and σ is the Warburg coefficient. -0.5 The linear fitting of is obtained.

[0025] Re-test other batches of batteries for logic correction and data learning to obtain battery aging data, and make real-time predictions on battery life based on battery usage parameters.

[0026] The content of the present invention is described in detail with a specific lithium battery; Example

[0027] 1) Take 9 lithium-ion batteries manufactured under the same conditions and test their initial capacity, internal resistance, voltage and electrochemical impedance spectroscopy (EIS) at 25°C; 2) Cycle the lithium-ion battery under three or more temperature conditions of 45°C / 65°C / 80°C, and the cycle current is the nominal current of the battery; 3) Test EIS, capacity, and internal resistance after 100, 500, 1000, and 2000 cycles and when the battery reaches 80% of its initial capacity (SOH retention rate is 80%); 4) Another batch of 9 batteries were placed at three or more temperature conditions of 45℃ / 65℃ / 80℃ for 0h, 500h, 1000h, 1500h, 2000h; 5) During the test, the DC internal resistance (DCR), capacity, and internal resistance (ACR) of the battery are tested; the above test data logic is fitted separately, and combined and fitted according to different battery usage conditions.

[0028] 6) Confirm the capacity attenuation degree and impedance increase according to the operating conditions to confirm the end-of-life node. Generally, the battery life is terminated by default when the capacity decays to 70% of the nominal capacity. Example

[0029] 1) Take 12 lithium-ion batteries manufactured under the same conditions and test their initial capacity, internal resistance, voltage and electrochemical impedance spectroscopy (EIS) at 25°C; 2) Cycle the lithium-ion battery under three or more temperature conditions of 5℃ / 45℃ / 60℃ / 85℃, and the cycle current is the nominal current of the battery; 3) Test cycles of 100, 500, 1000, and 2000 times. And EIS, capacity, and internal resistance when the battery is cycled to 80% of its initial capacity (capacity retention rate SOH is 80%); 4) Twelve batteries from the same batch were placed at 5℃ / 45℃ / 65℃ / 85℃ for 0h, 500h, 1000h, 1500h, 2000h at three or more temperature conditions; during this period, the DC internal resistance (DCR), capacity, and internal resistance (ACR) of the batteries were tested; 5) The above test data logics are fitted separately and combined according to different battery usage conditions.

[0030] 6) Confirm the capacity attenuation degree and impedance increase according to the operating conditions to confirm the end-of-life node. Generally, the battery life is terminated by default when the capacity decays to 70% of the nominal capacity.

[0031] The test of the present invention has high repeatability, small amount of calculation and high accuracy. Compared with the existing accelerated aging method, the essential characteristics and creativity of the present invention are embodied in: 1. Through the method of the present invention, a method for predicting the life of a lithium-ion battery is obtained. The method has the advantages of low cost, simple equipment requirements, and short preparation steps, and has the ability to correctly evaluate the life of a long-life lithium-ion battery.

[0032] 2. The present invention accelerates the actual life evaluation of long-life lithium-ion batteries by creatively using different attenuation rates of batteries under different temperature conditions, provides a practical application evaluation method, and provides data support for the maintenance cycle in product terminal applications. Temperature selection, including the upper and lower temperature limits of theoretical design, is set according to the size of the electrochemical window of the battery design through a combination of theoretical design and actual measurement.

[0033] 3. In the present invention, continuous charge and discharge cycle and static aging are used for evaluation. The actual battery usage conditions can be weighted and calculated, including modes of different usage conditions, with high compatibility.

[0034] 4. The present invention uses repeated testing to modify and learn the current logic. The accuracy gradually converges and improves. In addition, the measured data can be predicted with higher accuracy, which is different from pure logical operations or data processing.

[0035] The present invention has the advantages of simple process steps, short process, low equipment requirements for the entire route, good data stability, reduced energy consumption, and important industrial promotion and application value in preparing lithium iron phosphate positive electrode materials.

Claims

1. A method for predicting and evaluating the life of a long-life lithium-ion battery, characterized in that: The specific steps are as follows: (1) Take out 3n lithium-ion batteries manufactured under the same conditions and test the initial capacity, internal resistance, voltage and electrochemical impedance spectroscopy (EIS) at 25°C; (2) The lithium-ion battery is cycled under the temperature conditions of T1 / T2 / T3 / ··· / Tn, where n is greater than 3 and the cycle current is the nominal current of the battery. The test cycles are 100, 500, 1000 and 2000 times, and the EIS impedance, capacity and internal resistance are measured when the battery reaches 80% of its initial capacity; (3) Take 3n batteries of the same model as (1) and place them under the temperature conditions of T1 / T2 / T3 / ·· / Tn, where n is greater than 3, and the placing time is 0h, 500h, 1000h, 1500h and 2000h. During this period, measure the DC internal resistance (DCR), capacity and internal resistance (ACR) of the battery; (4) Add the logical formula for fitting to the data in steps (2) and (3), merge and fit according to the usage conditions of different batteries. According to the type and design of the battery, the values ​​of R0 to R4 in the EIS four-circuit method that change with degradation are most different. The four-circuit method is used to express R0: battery internal resistance (SEI resistance), R1: positive electrode internal resistance; R2: negative electrode internal resistance; R3: electrolyte diffusion resistance; R0 / R1 / R2 / R3, Warburg coefficient (σ) represents the slope of the fitting line, ω represents the angular frequency in the low frequency region, Li + The diffusion coefficient is closely related to the Warburg coefficient (σ) [191] , the fitting calculation formula is as follows: Z' = Re + Rct + σω −0.5 (1) D Li + = R 2 T 2 / 2A 2 n 4 F 4 C 2 σ 2 (2) Where R is the ideal gas constant, T is the absolute temperature, A is the active area of ​​the electrode, n is the number of electrons lost or reduced in each molecule, F is the Faraday constant, C is the molar concentration of lithium ions intercalated in the LFP particles, and σ is the Warburg coefficient. -0.5 The linear fitting of is obtained; Before performing data learning calculations (Formula 1 and Formula 2), take batteries of the same model and repeat the test according to steps (1) to (4); The capacity attenuation degree and impedance increase are confirmed according to the operating conditions to confirm the end-of-life node. The battery life is terminated by default when the capacity decays to 70% of the nominal capacity.

2. The method for predicting and evaluating the life of a long-life lithium-ion battery according to claim 1, wherein The batteries manufactured under the same manufacturing conditions described in step (1) are products of the same structure manufactured using the same or similar equipment, and use the same positive electrode material, negative electrode material, electrolyte, foil and corresponding auxiliary materials.

3. The method for predicting and evaluating the life of a long-life lithium-ion battery according to claim 1, characterized in that: In step (2), the selection of T1 / T2 / T3 / ··· / Tn is segmented according to the battery design temperature range including the upper limit, lower limit and usage conditions of the battery design temperature.

4. The method for predicting and evaluating the life of a long-life lithium-ion battery according to claim 1, characterized in that: The current in step (2) is selected according to the designed nominal current cycle of the battery and is adjusted according to the current actually used by the battery.

5. The method for predicting and evaluating the life of a long-life lithium-ion battery according to claim 1, wherein The test formula of the DC internal resistance (DCR) in step (3) is DCR=(U1-U2) / (I2-I1), where U1 is the end voltage of step (2), U2 is the end voltage of step (3), I1 is the discharge current of step (2), and I2 is the discharge current of step (3).

6. The method for predicting and evaluating the life of a long-life lithium-ion battery according to claim 1, characterized in that: Use the fitting logic described in step (4) of the battery capacity and impedance data calculation to fit the decay logic formula.

7. The method for predicting and evaluating the life of a long-life lithium-ion battery according to claim 1, characterized in that: The repetitive operation of step (5) is to modify and learn the impedance data calculated by the logic of step (4), and summarize the change trend of the battery system and fit the formula according to the growth law of the four resistances R0, R1, R2, and R3, the capacity decay rate, the charge and discharge time, the battery static time and other parameters.

Images