Learning and Behavior Predictive Control for Robots Based on Cloud Computing

Abstract

In this paper, the learning and behavior predictive control based on cloud computing is proposed for efficiently planning autonomous real time prespecifled trajectory tracking and obstacle avoidance control for an omnidirectional wheeled robot using fuzzy inference algorithm. The autonomous trajectory tracking control includes dynamic simulation according to object surface and depth measurement. The robot is equipped with three independent driven omnidirectional wheels and six ultrasonic sensors. The Jacobian between Cartesian space with respect to the joint space is setup for ellipse motion planning so that it not only can autonomously follow the prespecifled trajectory tracking but also avoid obstacles. An architecture is setup to split computation between the remote cloud and the robots so that the robots can interact with the computing cloud. Given this robot/cloud architecture, the stability of the closed loop control system using the predictive control algorithm is guaranteed with satisfactory tracking performance on the cloud during a periodically updated preprocessing phase, and manipulation queries on the robots given changes in the workspace can achieve real time trajectory tracking and obstacle avoidance. Finally, experiments are given to validate the path tracking performance and computational efficiency.