Research On Coordinated Control Of Multiple Mobile Robots Based On Artificial Potential Field Theory

Coordinated control of multiple mobile robots is becoming a hot topic in robot field recent years. Among the methods which are used to research on this topic, artificial potential field method (APF) is applied most widely till now. This method is the extension of the potential field concept in physics, which has the advantage of compact computation and nice real time property. Trough setting potential field contacts among robots or between robots and environment, APF can solve the problems of local path planning without collision and coordinated group formation of multiple mobile robots. The research of this thesis has divided the problem of coordinated control of multiple mobile robots into two sub problems: signal robot control and robots group control.First, the disadvantages of traditional APF are presented through theory and simulation. The obstacle avoidance of a single mobile robot is solved by using new APF based on relative position thought, which can avoid the problem of global minimum and unforced collision. Moreover, improvement is made in this new APF method considering some other drawbacks of traditional APF.Then in the group level, flocking, which is the basic model of group navigation, is discussed at first. Based on this model, through introducing the thought of relative position of single robot, new APF model for multiple mobile robots is made, and the choices of parameters of internal artificial potential function in this new model are proved as a focus. The problem of unforced collision and inter-robot collision can be reduced with proper parameters. Then an extension of APF method - stream function is researched. On the precondition of solving the obstacle avoidance of a single mobile robot using this function, swarming control model of multiple mobile robots based on dynamic connection rules is introduced to improve the performance of coordinated control with analyzing all sorts of connection rules. Finally, simulations are presented about the above coordinated control models of single mobile robot and multiple mobile robots. The results are analyzed and corresponding problems are introduced. The results demonstrate the correctness of the control law of coordinated control model and reveal the validity and the availability of the new APF model and the swarm model.