Electric vehicle SOC estimation method based on crossing-side suppression width learning

Abstract

The invention discloses an electric vehicle SOC estimation method based on spanning-side suppression width learning. The method comprises the steps that a training data set is acquired; constructing a crossing-side suppression width learning system; training the crossing-side suppression width learning system through the training data set; and estimating the SOC of the electric vehicle through the trained crossing-side suppression width learning system. According to the method, the SOC of the battery in different working states can be effectively estimated only by considering external characteristics of the battery and learning related sample data of the battery due to good nonlinear mapping capability and learning capability of the spanning-side suppression width learning system, so that the SOC estimation efficiency is improved. The battery SOC estimation method provided by the invention can avoid the problems existing in the SOC estimation method of a Kalman filtering method, and is more convenient and practical than a mechanism modeling method based on a physical model.

Description

Technical field [0001] The invention relates to the technical field of electric vehicles, in particular to an electric vehicle SOC estimation method based on cross-side suppression width learning. Background technique [0002] With the rapid development of the electric vehicle (Electric Vehicle, EV) market, the research on the core key technologies related to automobile batteries, which seriously affects and restricts the power performance of electric vehicles, has become a hot issue. Among them, the battery state of charge (State of Charge, SOC) is an important parameter in the battery management system, which can be used to characterize the remaining energy of the car battery and its working state, so as to reasonably arrange charging time and other related matters, so as to ensure the reliable and safe operation of electric vehicles. Therefore, the development of electric vehicles The prediction and estimation of battery SOC and its related fault diagnosis technology have...

AI Technical Summary

Problems solved by technology

However, although deep learning is an effective learning method, due to its large number of parameters to be optimized, it usually takes a lot of time and machine resources to optimize, which limits the neural network based on deep learning to a certain extent. system application

Images