Vision Based Autonomous Lateral and Longitudinal Control System

Abstract

Automation combined with the increasing market penetration of on-line communication, navigation, and advanced driver assistance systems will ultimately result in Intelligent Vehicle Highway Systems. Advanced Vehicle Control Systems (AVCS) is a key technology for Intelligent Transportation System (ITS) and Intelligent Vehicle Control System (IVCS). The unmanned control of the steering wheel is one of the most important challenges faced by the researchers. This paper proposes 1D sensor calibration algorithm which makes the vehicle to identify the track independent of various lighting conditions. Pre-processing algorithm helps in finding the position and orientation of the vehicle in the track. A novel control architecture for automatic steering, acceleration and braking control of a self guided vehicle has been proposed which consists of cascaded Kalman filter with PID steer control algorithm. It provides smooth tracking over the entire track. Adaptive speed control based on speed difference and acceleration control algorithm based on steer error are used. These algorithms make the vehicle to optimally manoeuvre the track. The experimental results show that the combination of Kalman filter with PID for lateral control reduces trajectory error to a minimum level and adaptive speed control algorithm for longitudinal control provides smooth speed over the entire track.