【作者】 吴绍斌； 李剑峰； 高利； 王刘安；

【Author】 Wu Shao-bin~1,Li Jian-fen~2,Gao Li~1,Wang Liu-an~1 (1.Beijing Institute of Technology,Beijing 100081; 2.China North Vehicle Research Institute,Beijing 100072)

【机构】 北京理工大学机械与车辆工程学院； 中国北方车辆研究所；

【摘要】 基于虚拟驾驶仿真平台开发了车辆稳定性神经网络控制器.以横摆角速度为控制目标,利用NARMA-L2神经网络估计模型辨识车辆动力学模型,通过调整神经网络对象模型设计了神经网络控制器,该控制器通过产生横摆力矩迫使车辆横摆角速度跟踪理想响应.在虚拟驾驶仿真平台上进行了操纵稳定性的虚拟试验,实现了驾驶员在环操纵控制仿真.仿真实验结果表明,设计的控制器能够有效改善车辆的稳定性.更多还原

【Abstract】 The paper demonstrates the implementation of a neural network controller based on the virtual driving simulation platform for the purpose of improving vehicle stability.The NARMA-L2 neural network approximate model was used to identify vehicle dynamics system,through rearranging the neural network object the neural network controller was developed to generate additional yaw moment and make the yaw rate tracking the desired trajectory.In order to verify the neural network controller the virtual experiments of vehicle handling stability were carried out on the virtual driving simulation platform,realizing human-in-loop simulation of driving maneuver.The simulation results show that the neural network controller can improve vehicle handling stability significantly.更多还原