Globally converging MIMO optimal controller for adaptive manipulation of mobile robots with redundant arms

In this paper, the stability of mobile robots during object manipulation with redundant arms is investigated. A new fast-converging MIMO control algorithm, called the Circles of INitialization (COIN), is introduced to calculate globally optimal postures of redundant manipulators. The algorithm is employed in real-time to prevent mobile robots tip-over during dynamic eccentric manipulation. This is achieved with tip-load measurement by adapting the position of the base relative to the line of action, and recalculating the global joint posture that offsets the risk of toppling. This closed-loop control scheme to maximize the arm´s payload and stabilize the robot´s posture is applied to serial redundant manipulators with n-joints.