The problem of coordinated control of multiple autonomous vehicles is a field of increasing research interests, due to its application in mapping, rescue mission, surveillance, search, reconnaissance, air traffic management systems, etc. A particular problem considered in the literature is formation control. Formation control is a type of coordinated behavior where mobile robots are required to autonomously configure into a desired spatial pattern. Formation maneuvering refers to the special case where the desired formation is not static, but moves in space as a virtual rigid body according to a predefined trajectory. Over the past decades many researchers have worked on the formation control with differences regarding the type of agent’s kinematics/ dynamics, varieties of control strategies, and the type of the task demanded.
In our research we present a solution to the formation maneuvering problem for multiple nonholonomic wheeled mobile robots. Our solution is based upon the graph that models the coordination among the robots being a spanning tree. Our control law ensures, in the least square sense, that the robots globally asymptotically acquire a given planar formation while the formation as a whole globally asymptotically tracks a desired trajectory. The control law is to first designed at the kinematic level and then extended to the dynamic level. In the latter, we consider that parametric uncertainty exists in the equations of motion. These uncertainties are accounted for using an adaptive control scheme. The proposed formation maneuvering control is demonstrated by numerical simulations of three, four, and five vehicles.
Current work: Experimental validation of the formation maneuvering control algorithm using the TurtleBot platform.
