Method for online estimating capacity and health state of battery pack based on random charging information

Abstract

The invention relates to the technical field of electric vehicles, and particularly discloses a method for online estimating the capacity and health state of a battery pack based on random charging information, which comprises the following steps of: preliminarily estimating the capacity of the battery pack: selecting a charging section and utilizing an ampere-hour integral to estimate a model; temperature-based battery capacity correction: temperature data is corrected based on fuzzy Kalman filtering optimization algorithm processing, and because relatively large noise is inevitably generated after preliminary estimation, a fuzzy control algorithm is adopted to control observation noise of a Kalman filtering algorithm in real time, so that an estimation result is converged to a real trend; estimating and predicting the state of health of the battery: establishing an Arrhenius model based on the capacity attenuation of the battery pack to predict the service life of the battery; according to the method, the defect of inaccurate estimation of real-time measurement noise of a traditional Kalman filtering algorithm is effectively overcome, and the measurement noise is corrected in real time; the method is suitable for AI adaptive parameter adjustment models of different new energy heavy truck models, and is applied to state-of-health evaluation and life prediction of batteries.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicles, and in particular to a method for online estimating battery capacity and health status based on random charging information. Background Art

[0002] In the field of heavy-duty commercial vehicles, profitability is the top priority for customers, and its core lies in creating greater value for customers. So how to continuously improve the profitability of heavy-duty trucks? Customers' focus is on reducing the vehicle's TCO (Total Cost of Ownership). The TCO of heavy-duty trucks is mainly composed of three parts: fixed costs (insurance, driver labor costs), variable costs (fuel consumption, maintenance, tires) and depreciation. Under the new energy trend of heavy-duty trucks, the electricity used by electric heavy-duty trucks is environmentally friendly, stable and cheap. Compared with traditional heavy-duty trucks, they have more advantages in fuel costs; in addition, the maintenance cycle is long, the maintenance items are few, and the maintenance cost will also be lower. In order to solve the dilemma of high sensitivity of TCO of heavy-duty trucks in the new energy era, the service life of batteries has gradually become the key.

[0003] Aging affects the energy and power output capacity of the battery, which in turn affects the economy and power of the vehicle. How to manage the life of power batteries and evaluate their assets has become a key issue in the electric vehicle industry. In order to better detect and evaluate the performance of batteries after application in electric heavy trucks, it is crucial to find an effective battery pack life estimation solution.

[0004] There are two methods for estimating battery life in the existing technology, including capacity estimation and internal resistance estimation. Internal resistance estimation can be measured by electrochemical impedance spectroscopy (EIS), but due to the low recording frequency, it is difficult to apply to real-time control; while calculations through direct voltage response, battery model-based, neural network model and other methods have the problem of difficulty in achieving both accuracy and speed, and the internal resistance is complexly affected by temperature and SOC. In-vehicle applications, capacity is often used as an indicator for evaluating the state of health (SOH) of the battery. However, based on ordinary ampere-hour integration or empirical models, the accuracy is generally reduced due to limitations on the accuracy of state of charge (SOC) estimation and open-loop processing.

[0005] Therefore, there is an urgent need to design a method for online estimation of battery pack capacity and health status based on random charging information to solve the problem of low online capacity accuracy and difficult life prediction of existing heavy-duty commercial vehicle power batteries based on ordinary ampere-hour integration or empirical models. Summary of the invention

[0006] In view of the problems existing in the prior art, the purpose of the present invention is to provide a method for accurately correcting capacity based on random charging data fragments using an optimization algorithm combining Kalman filtering and fuzzy control, which can be used to develop an Al adaptive parameter adjustment algorithm and is conveniently applied to the health status evaluation and life prediction of batteries of different models.

[0007] The technical solution adopted by the present invention to solve the technical problem is: a method for online estimating the capacity and health status of a battery pack based on random charging information, comprising the following steps: S1. Preliminary estimation of battery pack capacity: Select the charging segment and use the ampere-hour integral estimation model; S2. Battery capacity correction based on temperature: The correction of temperature data is based on the optimization algorithm of fuzzy Kalman filter; S3. Battery health status estimation and prediction: Based on the battery pack capacity decay, an Arrhenius model is established to predict the battery life.

[0008] Preferably, the battery pack in step S1 includes a plurality of cells, and the ampere-hour integral method is used to estimate the charging time of any single cell, and the ampere-hour integral estimation model is as follows: ; Where: Cap is the cell capacity calculated in the charging segment, t1 represents the start time of the selected charging segment, t2 represents the end time of the selected charging segment, I(t) represents the battery current at time t, SOC1 and SOC2 are the SOC values ​​corresponding to time t1 and t2 respectively.

[0009] Preferably, the temperature characteristic of the battery in the battery capacity correction in step S2 is that in a low temperature environment, the discharge capacity of the battery is significantly reduced; in a high temperature environment, the discharge capacity of the battery is significantly increased; considering the influence of temperature on the battery capacity estimation result, the estimation result is corrected according to the temperature data to make the battery capacity estimation result more consistent with the actual change of the power battery, therefore 25°C is set as the standard temperature for temperature correction, and correction is made according to the temperature data to obtain the standard capacity.

[0010] Preferably, the battery capacity changes slowly, so when establishing the EKF state equation, it is considered that the capacity at time i+1 is equal to the capacity at time i, but there is system noise in the Kalman filter algorithm model, and the state equation of the battery capacity is: ; In the formula, x i represents the best estimated battery capacity, w i represents the system noise; The battery capacity output equation is: ; In the formula, vi represents the measurement noise.

[0011] Preferably, the fuzzy control algorithm is a fuzzy logic that imitates the uncertainty concept judgment and reasoning thinking mode of the human brain to achieve fuzzy comprehensive judgment, which mainly includes three parts: fuzzification, fuzzy rule algorithm and defuzzification; the input parameters of the Kalman filter-fuzzy control algorithm include three values: SOC1, SOC1 and relative error of battery capacity observation; the fuzzy rule algorithm is based on the selected charging segment Different weight factors are selected for different locations ; The battery capacity based on temperature correction is optimized using an optimization algorithm based on fuzzy Kalman filtering to obtain a high-precision battery capacity.

[0012] Preferably, the battery health state in step S3 is SOH, which is defined as the ratio of the current capacity of the battery to the initial capacity. The SOH of the battery pack is calculated by selecting the cell with the minimum capacity in the battery pack as the standard. Then, the SOH of the battery pack is calculated by the following formula: ; Where Cap(i) is the current capacity value of the selected battery cell, and Capinit(i) is the initial capacity value of the selected battery cell; Based on the accumulation of random charging data, a data set between SOH and time n over a period of time is obtained, and an Arrhenius model is established based on the battery pack capacity decay: ; In the formula, is the SOH corresponding to time n; A is the pre-exponential factor; E a is the activation energy, which is independent of temperature and has the unit of ; T is absolute temperature, unit is K; R is gas molar constant, unit is .

[0013] Preferably, the SOH(i) and They are respectively used for battery health status evaluation and life prediction.

[0014] The present invention has the following beneficial effects: The method designed by the present invention for online estimation of battery capacity and health status based on random charging information is combined with the Kalman algorithm of fuzzy control, which effectively solves the shortcomings of inaccurate real-time measurement noise estimation of the traditional Kalman filter algorithm, improves the influence of measurement noise on the estimation result as the working conditions change in the traditional Kalman filter algorithm, and can also perform real-time correction of the measurement noise.

[0015] The method designed by the present invention for online estimation of battery capacity and health status based on random charging information is combined with Kalman filtering and fuzzy control algorithm optimization, and is applied to online correction of temperature data to obtain better online capacity estimation.

[0016] The method designed by the present invention for online estimation of battery pack capacity and health status based on random charging information is based on battery capacity attenuation data corrected by temperature optimization. The method of the present invention is used to develop an Al adaptive parameter adjustment algorithm suitable for different new energy heavy-duty truck models, which is applied to battery health status evaluation and life prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Flow chart of a method for online estimation of battery pack capacity and health status based on random charging information.

[0018] Figure 2 It is a calculation model diagram of a method for online estimation of battery pack capacity and health status based on random charging information. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely further describe the technical solutions in the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0020] like Figure 1-Figure 2 As shown, the method for online estimating battery capacity and health status based on random charging information includes the following steps: 1. Preliminary estimation of battery pack capacity: Select the charging segment and use the ampere-hour integral estimation model.

[0021] The battery pack consists of multiple cells. The ampere-hour integral method is used to estimate the charging stage of any single cell. The ampere-hour integral estimation model is as follows: ; Where: Cap is the cell capacity calculated in the charging segment, t1 represents the start time of the selected charging segment, t2 represents the end time of the selected charging segment, I(t) represents the battery current at time t, SOC1 and SOC2 are the SOC values ​​corresponding to time t1 and t2 respectively.

[0022] 2. Battery capacity correction based on temperature: The correction of temperature data is based on the optimization algorithm of fuzzy Kalman filter.

[0023] According to the temperature characteristics of the battery, the discharge capacity of the battery is significantly reduced in a low temperature environment, and the discharge capacity of the battery is significantly increased in a high temperature environment. Taking into account the impact of temperature on the battery capacity estimation results, the estimation results are corrected according to the temperature data to make the battery capacity estimation results more consistent with the actual changes of the power battery. Therefore, 25°C is set as the standard temperature for temperature correction, and correction is made according to the temperature data to obtain the standard capacity.

[0024] Since large noise will inevitably be generated after the initial estimation, a fuzzy control algorithm is used to control the observation noise of the Kalman filter algorithm in real time so that the estimation result converges to the real trend.

[0025] Considering that the battery capacity changes slowly, when establishing the EKF state equation, it is assumed that the capacity at time i+1 is equal to the capacity at time i. However, there is system noise in the Kalman filter algorithm model, and the state equation of the battery capacity is: ; In the formula, x i represents the best estimated battery capacity, w i Represents system noise.

[0026] Since the capacity change of the battery between charging intervals is small, the system noise w meets the filtering requirements by a small value; The battery capacity output equation is: ; In the formula, v i represents the measurement noise.

[0027] The fuzzy control algorithm is a fuzzy logic that imitates the uncertainty concept judgment and reasoning thinking mode of the human brain to achieve fuzzy comprehensive judgment. It mainly includes three parts: fuzzification, fuzzy rule algorithm and defuzzification; the input parameters of the Kalman filter-fuzzy control algorithm include three values: SOC1, SOC1 and relative error of battery capacity observation; the fuzzy rule algorithm is based on the selected charging segment Different weight factors are selected for different locations The battery capacity based on temperature correction is optimized using an optimization algorithm based on fuzzy Kalman filtering to obtain a high-precision battery capacity.

[0028] 3. Battery health status estimation and prediction: Based on the battery pack capacity decay, the Arrhenius model is established to predict the battery life.

[0029] The battery health state is SOH, which is defined as the ratio of the current capacity of the battery to the initial capacity. The SOH of the battery pack is calculated based on the minimum cell in the battery pack. The SOH of the battery pack is calculated by the following formula: ; Where Cap(i) is the current capacity value of the selected battery cell, and Capinit(i) is the initial capacity value of the selected battery cell.

[0030] Based on the accumulation of random charging data, a data set between SOH and time n within a period of time is obtained, and an Arrhenius model is established based on the battery pack capacity decay. Specifically: ; In the formula, is the SOH corresponding to time n; A is the pre-exponential factor; E a is the activation energy, which is independent of temperature and has the unit of ; T is absolute temperature, unit is K; R is gas molar constant, unit is .

[0031] The above SOH(i) and They are respectively used for battery health status evaluation and life prediction. In addition, using this algorithm as the original model, an AI adaptive parameter adjustment model can be developed for extended application based on data accumulated during the charging process of different vehicle models, which can be conveniently applied to battery health status evaluation and life prediction of different vehicle models.

[0032] The present invention is not limited to the above-mentioned implementation modes, and anyone should be aware of the structural changes made under the enlightenment of the present invention, and all those having the same or similar technical solutions as the present invention fall within the protection scope of the present invention.

[0033] The techniques, shapes, and structural parts not described in detail in the present invention are all well-known techniques.

Claims

1. A method for online estimation of battery capacity and health status based on random charging information, characterized in that: The following steps are involved: S1. Preliminary estimation of battery pack capacity: Select the charging segment and use the ampere-hour integral estimation model; S2. Battery capacity correction based on temperature: The correction of temperature data is based on the optimization algorithm of fuzzy Kalman filter; S3. Battery health status estimation and prediction: Based on the battery pack capacity decay, an Arrhenius model is established to predict the battery life.

2. The method for online estimating battery capacity and health status based on random charging information according to claim 1, characterized in that: The battery pack in S1 includes multiple cells. The ampere-hour integral method is used to estimate the charging stage of any single cell. The ampere-hour integral estimation model is as follows: ； Where: Cap is the cell capacity calculated in the charging segment, t1 represents the start time of the selected charging segment, t2 represents the end time of the selected charging segment, I(t) represents the battery current at time t, SOC1 and SOC2 are the SOC values ​​corresponding to time t1 and t2 respectively.

3. The method for online estimating battery capacity and health status based on random charging information according to claim 1, characterized in that: The temperature characteristic of the battery in the battery capacity correction in S2 is that the discharge capacity of the battery decreases in a low temperature environment; and the discharge capacity of the battery increases in a high temperature environment; the estimation result is corrected according to the temperature data so that the battery capacity estimation result conforms to the actual change of the power battery, 25°C is set as the standard temperature for temperature correction, and correction is performed according to the temperature data to obtain the standard capacity.

4. The method for online estimating battery capacity and health status based on random charging information according to claim 3, characterized in that: The state equation of the battery capacity, when establishing the EKF state equation, the capacity at time i+1 is equal to the capacity at time i. Based on the system noise in the Kalman filter algorithm model, the state equation of the battery capacity is: ； In the formula, x i represents the best estimated battery capacity, w i represents the system noise; The battery capacity output equation is: ； In the formula, v i represents the measurement noise.

5. The method for online estimating battery capacity and health status based on random charging information according to claim 2, characterized in that: The optimization algorithm of the fuzzy Kalman filter in S2 includes a fuzzy control algorithm. The fuzzy control algorithm is a fuzzy logic that imitates the uncertainty concept judgment and reasoning thinking mode of the human brain to achieve fuzzy comprehensive judgment. It mainly includes three parts: fuzzification, fuzzy rule algorithm and defuzzification; the input parameters of the Kalman filter-fuzzy control algorithm include three values: SOC1, SOC1 and the relative error of battery capacity observation; the fuzzy rule algorithm is based on the selected charging segment Different weight factors are selected for different locations ; The battery capacity based on temperature correction is optimized using an optimization algorithm based on fuzzy Kalman filtering to obtain a high-precision battery capacity.

6. The method for online estimating battery capacity and health status based on random charging information according to claim 1, characterized in that: The battery health status in S3 is SOH, which is defined as the ratio of the current capacity of the battery to the initial capacity. The SOH of the battery pack is calculated by selecting the cell with the minimum capacity in the battery pack as the standard. Then the SOH of the battery pack is calculated by the following formula: ； Where Cap(i) is the current capacity value of the selected battery cell, and Capinit(i) is the initial capacity value of the selected battery cell; Based on the accumulation of random charging data, a data set between SOH and time n over a period of time is obtained, and an Arrhenius model is established based on the battery pack capacity decay: ； In the formula, is the SOH corresponding to time n; A is the pre-exponential factor; E a is the activation energy, which is independent of temperature and has the unit of ; T is absolute temperature, unit is K; R is gas molar constant, unit is .

7. The method for online estimating battery capacity and health status based on random charging information according to claim 6, characterized in that: The SOH(i) and They are respectively used for battery health status evaluation and life prediction.

Images