Formation Control Of Mobile Robots And Its Application In The Odor Source Tracking

Multiple autonomous robot system (MARS), which is the most challengingsubject in the research area of control and robot, has attracted many researchers’attention in recent years. Each robot in a MARS is designed to accomplish the desiredtasks by resembling some social behaviors of animals. Most tasks in which the MARSis applied are faced with a common key issue, i.e., formation control. The applicationof multiple robots rather than single robot in source tracking could be effective inexpanding the sensing zone. Thus the accuracy and efficiency of plume tracking couldbe greatly increased.This thesis focuses on the study of formation control and the application of theMARS in odor-source tracking. The main contribution is presented as follows:Firstly, a brief overview of MARS formation control is introduced. Additionally,the research background, development prospect and major problems are alsoprovided.Secondly, a novel distributed MICE (Multi-robot for Intelligent CooperativeExploration) platform consisting of a group of12robots is designed. The mechanismmodel, sensing and control units of the platform are explained.Thirdly, simulation on the classical Leader-Follower control algorithm is carriedout on the VC platform. Based upon the simulation, a successful result is obtained onthe MICE platform.Fourthly, given the distributed and non-holonomic features of MARS, a modifiedformation control algorithm for the simplified chain-form model is proposed.VC/Matlab based simulation results on the simplified model validate the effectivenessof the algorithm.Lastly, the formation control algorithm on the plume model proposed by Farrellis studied. The simulation of the algorithm on the odor-source tracking is successfullyimplemented.