Intelligent Vehicle's Path Tracking Based on Fuzzy Control

Autopilot vehicle is an important part of intelligent transportation systems. The objective is to develop the driver assistance systems on highway and urban road, to help or even to replace the driver, which may reduce traffic accidents and improve the efficiency of traffic system. A method based on machine vision and fuzzy control is proposed to realize intelligent vehicles' autopilot. It uses the CMOS sensor as its path recognition device to draw its lane centerline through image analysis. Taking the feedback speed as the additional input, the study forms the closed-loop control and establishes one graduation fuzzy controller which controls vehicle direction with two fuzzy controller combinations and replaces traditional PID control vehicle speed by fuzzy control. Compare with the conventional PID algorithm and the fuzzy control algorithm, the improved fuzzy control algorithm ensures a high speed and steady running of intelligent vehicle with smaller over modulation in corner.

摘要车辆无人驾驶是智能交通系统的一个重要部分，其目标是开发在高速公路和城市道路环境下的辅助驾驶系统，旨在帮助乃至取代驾驶员，实现车辆自动控制和自动驾驶，减少交通事故发生，提高道路交通系统的效率，因此提出了一种基于机器视觉和模糊控制实现智能车辆自主行驶的方法。该方法以CMOS摄像头为路径识别传感器，通过图像分析提取车道中心线，并引入速度反馈，形成闭环控制，建立一个由两个模糊控制器组成的分级模糊控制器控制车辆转向，并使用模糊控制代替传统的PID速度控制来控制速度。和常规的PID算法及模糊控制算法相比，改进的模糊控制算法使智能车在道路上更快速、平稳地运行，并且在转弯处的超调更小。