Lithium ion power battery SOC estimation method based on neural network optimization EKF

Abstract

The invention provides a lithium ion power battery SOC estimation method based on neural network optimization EKF. On one hand, a Thevenin equivalent circuit model is built according to nonlinear voltage characteristics of a lithium ion battery, model parameters are determined on the basis of experiments of different SOC points and charge-discharge directions, a state equation and an observation equation are obtained based on a lithium ion battery model, and a calculation process of an extended Kalman filtering-based SOC estimation algorithm is designed; and on the other hand, an error prediction model is built based on a BP neural network, and a measurement noise covariance is corrected in real time in a filtering process, so that a state estimation error introduced due to the fact that amodel error is relatively large and system noise is assumed to be Gaussian white noise is eliminated. According to the method, an SOC estimation result of the EKF is compensated based on various modeling errors; the superiority of combination of the BP neural network and the EKF algorithm is proved; the maximum estimation error is within 0.25%; and the method has relatively high engineering application values.

Description

technical field [0001] The invention relates to the technical field of battery detection, in particular to a method for estimating the SOC of a lithium-ion power battery based on a neural network optimized EKF. Background technique [0002] Due to the advantages of zero emission, zero pollution, and high energy utilization rate, electric vehicles have great potential for development in new energy vehicles. The battery management system is an important part of electric vehicles, and the power battery state of charge (SOC) estimation It is also one of the key technologies of the battery management system. SOC estimation involves power battery charge and discharge control and optimal management of electric vehicles, which directly affects the service life of the power battery and the performance of the power system. If the SOC of the battery cannot be accurately estimated, the battery will be overcharged or overdischarged, resulting in Therefore, accurate estimation of SOC is ...

AI Technical Summary

Problems solved by technology

For example, the battery model error can be established as a piecewise linear function of SOC and current, and the measurement noise of the EKF can be corrected according to the calculated model error, which improves the estimation accuracy of SOC to a certain extent, but the established measurement noise correction model It is qualitative and needs to be obtained based on experience, not universal

Images