基于机器视觉的车辆自动驾驶模糊控制设计

Fuzzy Controller Design of Automatic Driving for Automobiles Based on Machine Vision

针对车辆动力学控制系统所具有的强非线性特点提出了基于机器视觉的车辆自动驾驶模糊控制方案。采用Michael等人提出的车辆系统动力学模型，通过模糊控制规则的量化划分对车辆在道路上的运动进行了仿真。仿真的结果显示，本方案可以很好地解决空旷道路上的车辆自动驾驶问题，并且该控制方法可以保证车辆快速准确地在道路上安全高速行驶，具有很好的鲁棒性。

A fuzzy controller of automatic driving based on machine vision was designed according to the strong nonlinear of vehicle dynamics system. A reasonable vehicle systematic dynamical model was adopted, fuzzy control rules were quantized and subdivided. The simulation of the moving trajectory of the vehicle on road showed a preferable solution to the automatic driving on an empty road, furthermore, this control strategy can ensure vehicle run on various reference paths quickly, accurately and robustly.