Fuzzy PD Controller Based Path Tracking for Autonomous Vehicle under i-VICS

Abstract

The autonomous driving under the environment of Intelligent Vehicle-Infrastructure Cooperative Systems (i-VICS) is a brand-new technical route that can reduce the dependence of driverless autopilot on high-precision sensors to detect the path surroundings and intelligent calculations to make the operational decisions. Based on i-VICS, the most of data to support the autonomous driving is then provided from road-side units (RSUs) based on V2I communication. It is a highly reliable and real-time solution at low cost compared with the traditional approaches of autonomous driving. This paper discusses the control of path tracking as the basic testament induced from the newly developed technical route. The design of a path tracking algorithm based on an improved model to derive the lateral deviation via multiple assigned preview points on reference path is explored to reduce the impact of noises from reference path data. The strategy to select the optimal look-ahead distance used for deviation calculation is also analyzed. A fuzzy PD controller is then introduced to support the control of steering and the speed control based on path geometry is designed for the longitudinal control. The simulation and field tests results demonstrate a satisfactory performance finally.