A prediction model for vehicle sideslip angle based on neural network

Abstract

Sideslip angle is the most widely used attributes for measuring the vehicle side slipping. Predicting the trend of sideslip angle in advance is of great significance for sideslipping precaution. In this research, small-vehicle model was selected, took steering wheel angle, yaw rate, lateral acceleration and four wheel velocities into account, and then applied neural network to build a prediction model for the sideslip angle 0.5 second in advance. Through applying the model to predict the sideslip angle based on data stimulated by veDYNA, a vehicle dynamics stimulation software, and comparing to the observation of sideslip angle produced by veDYNA, it testified that the forecast model is highly accurate.