Storage battery state of charge estimation method based on self-adaptive unscented Kalman filtering

Abstract

The invention discloses a storage battery state of charge estimation method based on self-adaptive unscented Kalman filtering. The method comprises that: performance of a storage battery is tested via an HPPC test so that HPPC test data of the storage battery are acquired; parameter identification is performed on the basis of the acquired HPPC test data of the storage battery so that storage battery model parameters are acquired; and an SOC of the storage battery is estimated on the basis of the acquired storage battery model parameters via an AUKF algorithm. According to the storage battery state of charge estimation method based on the self-adaptive unscented Kalman filtering, defects in the prior art that manual amount of labor is large, real-time performance is poor, operation difficulty is high, test and calculation accuracy is low, etc. can be overcome so that advantages of being small in manual amount of labor, great in real-time performance, low in operation difficulty and high in test and calculation accuracy can be realized.

Description

technical field [0001] The invention relates to the technical field of automobile batteries, in particular to a storage battery state-of-charge estimation method based on an adaptive unscented Kalman filter. Background technique [0002] In recent years, the development of power batteries has been extremely rapid, but correspondingly, the development of battery management technology has lagged behind seriously, which has also led to battery management technology becoming an important factor restricting the development of electric vehicles. Due to the imperfect management technology, the power battery for electric vehicles is in a state of overcharge or overdischarge for a long time, and the battery performance gradually deteriorates with use, resulting in high battery cost. Therefore, accurate estimation of SOC is particularly important. However, SOC is not a physical quantity that can be directly measured. The battery itself is a closed electrochemical reaction. When electr...

AI Technical Summary

Problems solved by technology

[0009] (1) It is more reliable to obtain SOC by the discharge test method, but the discharge test method is a completely offline measurement method, and for electric vehicles, a spare battery pack is required, and the cost increases too much

However, the security time method will lead to the accumulation of errors over time, which will lead to inaccurate initial values ​​for the next estimate, and require a large amount of data to be stored

[0011] (3) The open circuit voltage method is easy to use, but it requires the battery pack to stand intermittently, and it takes a certain period of time to be offline, which cannot meet the real-time requirements of electric vehicles

[0012] ⑷Neural network method, the estimation is more accurate, but it needs a lot of training data and suitable training algorithm, the training data is not easy to obtain, the suitable training algorithm is not easy to find, and the neural network needs to store a large amount of data, which increases the hardware cost

[0013] (5) The Kalman filter algorithm is more suitable for the severe operating conditions of electric vehicles, but the traditional Kalman filter is only suitable for linear systems, and is not applicable to the strong nonlinearity presented during battery use

[0014] ⑹The extended Kalman filter method adopts the first-order Taylor technique approximation, which can be used to estimate the state of charge of the battery, but the extended Kalman filter uses a linearization method that approximates the nonlinear system to a linear time-varying system. Linearization errors are generated, and the extended Kalman filter method needs to solve the Jacobian matrix of the system. The solution algorithm is complex and the operation speed is slow, which is not conducive to hardware implementation

[0015] (7) The effective implementation of the extended Kalman filter algorithm must rely on the accurate establishment of the battery model. When the electric vehicle is running, the battery is accompanied by a violent chemical reaction, and the circuit parameters will also change with the use of the battery. This change will inevitably lead to inaccurate SOC estimation. accurate

[0016] In the process of realizing the present invention, the inventor found that there are at least defects in the prior art such as large amount of manual labor, poor real-time performance, difficult operation and low precision.

Images