Formation Control Based On Leader-followers For Multiple Mobile Robots

Recently, the coordination control of multi-robot systems has attracted the attention of thelarge number scholars in engineering and academia due to several advantages such as redundancy,concurrency, robustness and its broad application in such areas including target search, cooperativeobject transport task, rescue missions and so on. Focusing on the odometer positioning error offormation motion and the communication burden of neighbors’ upstairs consistency algorithm, themain contents and contributions can be summarized as follows:Firstly, this paper introduces the research background and significance of multiple robotformations and the application situation of robot formation at home and abroad. Aimed at thestudy content of this article，some basic knowledge about the coordinate transformation methodunder reference coordinate system, several typical method of formation coordination and graphtheory and matrix theory in consistency algorithm research are introduced. Differential drive robotand leader–follower’s style is chosen to research. Under this model, the three kinds of basicmovement with reuse characteristics and corresponding control algorithm are studied.Secondly, for the problem of odometer positioning error of formation motion, we proposed asimple and effective method for formation control of multiple mobile robots in corridor or likecorridor environments. The corridor environments are regard as the reference frame of the directionand distance., and then the formation control problems of robot formation on the2D plane areconverted into a series of problem on the1D plane to make the final position of it be the same to hisvirtual robot. Finally, this strategy is applied to the leader-follower’s formation, and the specificoperation algorithm of each robot is presented combining with three kinds of robot motion behaviorin the previous chapter.Finally, for the problem of the communication in consensus algorithm based on second-orderneighbors’ information, a consensus algorithm using the nearest second-order neighbors’information, in which only partial second-order neighbors’ information is used, is proposed. Then,we conducted the stable analysis to compare the proposed algorithm with the consensus algorithmsbased on the first-order neighbors’ information, and based on the second-order neighbors’information on the convergence, respectively. Finally, the three algorithms were used to multiplerobot formations with leader-followers to compare on the robustness.