基于径向基神经网络的EPS助力控制方法

Control Method of EPS Power-assisted System Based on Radial Basic Function Neural Network

回顾汽车助力转向系统的发展历程,并对电动助力转向(EPS)系统进行分析。在总结前人研究方法的基础上,探索出一种基于径向基(RBF)神经网络的EPS助力特性研究方法。通过径向基神经网络的学习和训练,对理想的助力特性曲线进行拟合和预测,得到非常理想的新曲线。学习的结果说明,径向基神经网络在学习和预测方面功能强大,应用此控制策略,对后续的研究提供了一个选择。

This paper reviews the development process of auto-power-assisted steering system, and analyzes the electric power steering （EPS） system. According to summarizing research methods of others, it probes out a new research method which is based on radial basis function nerve network（RBFNN）. Through the learning and training of RBFNN, the paper fits and predicts the EPS curves, and gets its very ideal curves. These results illuminate that the learning and predicting function of RBFNN is strong, this control strategy may be a new choice in following research.