Simulation of Obstacle Avoidance for Auto Guided Land Vehicle

Abstract

This research is concerned with designing and simulating an auto control system for a car provided with obstacle avoidance sensors. This car is able to pass through predefined path an around the detected obstacles, and then come back to the intended path. The IR sensor detects the existence of the obstacle through an assumed range of detection, while the visual sensor (camera) feeds back an image including the path that contains  an obstacle, which can be useful for determining the obstacle's length, speed, and direction. According to such information, the controller creates transient away point along the longitudinal axis of the obstacle which is the same as the transverse axis of the simulator path at an assumed distance from the obstacle. The robotic car will direct toward the transient point for avoiding the obstacle, which directly comeback into the original rout once when reaching the transient point. This strategy enables the car to move far away from the obstacle and then return it into planned path.