基于RBF神经网络调节的电动车驱动和再生制动滑模控制  

Sliding mode control for driving and regenerative braking of electric vehicle based on RBF neural network

作　　者：曹建波[1] 曹秉刚[1] 王军平[1] 许朋[1] 武小兰[1] 

机构地区：[1]西安交通大学机械工程学院,西安710049

出　　处：《吉林大学学报（工学版）》2009年第4期1019-1024,共6页Journal of Jilin University:Engineering and Technology Edition

基　　金：陕西省中小企业创新基金项目(06C26216100555);陕西省自然科学基金项目(SJ08E218)

摘　　要：针对电动车的驾驶模式多变和续驶里程短等主要问题,在建立控制系统主回路的基础上对电动车的驱动和再生制动过程进行了研究,设计了电动车用神经网络滑模控制器。该控制器包括两部分:一个RBF神经网络和一个滑模控制器,RBF神经网络对滑模控制器进行在线切换增益调节。实验结果表明,在车辆的驱动和再生制动过程中,神经网络滑模控制器与普通滑模控制器相比,其响应速度、稳态误差及鲁棒性明显改善,尤其在再生制动过程中,可以回收更多的能量,进一步延长了车辆的续驶里程,对节约能源很有意义。Based on building the main circuit of control system for the electric vehicle（EV）, its driving and regenerative braking processes were studied to alleviate its main problems such as variation of driving mode and short driving range. A mathematic model was established for the EV system and a sliding mode controller with a neural network（NN） was designed. The controller consists of a radial basis function（RBF） NN and a sliding mode controller（SMC）. The RBFNN is used to adjust adaptively the gain of SMC on-line. The experiment results show that comparing with the traditional SMC, the NNSMC is characterized by better performances in dynamic response, steady-state tracking accuracy and system robustness when either driving or braking. In the EV braking process, the designed system can recover energy more than traditional SMC and lengthen the driving range, being in favor of energy-saving.

关 键 词：自动控制技术 电动车 驱动控制 再生制动 神经网络 滑模控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]