Different control strategies for Mobile Robots

Abstract

In this research, the following (ego) vehicle was maintaining a safe relative distance by varying the velocity through different controllers to catch up with the lead vehicle. Two major sensors were used, a rotating Laser Distance Sensor (LDS) and an RGB camera sensor. The camera sensor operated as a secondary system and was used to improve the detection probability of the lead vehicle. A sensor fusion algorithm was used for localizing the ego vehicle which includes a detection clustering algorithm and a Kalman filter to estimate the relative distance between the two vehicles. The sensors were calibrated for the test conditions to obtain detections with in feasible limits. Different controllers such as Proportional, Proportional-Integral, and Model Predictive Control were implemented and validated both in simulation and experiment on the turtlebot3-burger robot.