论文部分内容阅读
以智能车为控制对象,为解决单一运动控制方法难以满足智能车控制准确度、稳定性等要求的问题,设计了一种基于模糊神经网络的智能车运动控制器。该控制器利用神经网络实现模糊推理,将智能车与预定轨迹之间的相对距离和方位信息模糊化后作为输入,并通过神经网络学习算法不断调整连接权值和模糊值中隶属函数的参数,提高了控制器的自适应能力,使智能车能够快速稳定地沿预定道路行驶。仿真和实验结果均表明,该控制器动态响应好、稳态误差小,能够满足智能车的运动控制要求。
In order to solve the problem that single motion control method can not meet the requirements of accuracy and stability of intelligent vehicle control, a smart vehicle motion controller based on fuzzy neural network is designed. The controller realizes fuzzy reasoning by using neural network, fuzzifies the relative distance and orientation information between the smart car and the scheduled track as input, and adjusts the parameters of the membership function in the connection weights and the fuzzy values through the neural network learning algorithm, Increased adaptive capacity of the controller to enable smart cars to travel quickly and steadily along a predetermined road. Simulation and experimental results show that the controller has a good dynamic response and a small steady-state error, which can meet the motion control requirements of smart cars.