Search

The surface-mounted permanent-magnet synchronous motor is a two-input, two-output nonlinear system. The multi-input, multi-output nature of the system has posed some specific challenges to various control methods. Recently, the robust output regulation problem of the system subject to a known neutrally stable exosystem was studied. The problem came down to a global robust stabilization problem of an augmented system composed of the original plant and an internal model. In this paper, we will further study the robust output regulation problem of the system subject to an unknown neutrally stable exosystem. Like in the case where the exosystem is known, the current problem can be solved by globally stabilizing an augmented system. But unlike in the case where the exosystem is known, the augmented system takes a much more complicated form because of uncertainty in the exosystem than the case where the exosystem is known. In particular, the dynamic uncertainty in the current augmented system contains linearly parameterized uncertainty, and hence is not input-to-state stable. By utilizing some dynamic coordinate transformation technique, and combining some robust control and adaptive control techniques, we will solve the problem via a recursive approach. Copyright © 2014 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, we study the cooperative global adaptive output regulation problem for a class of nonlinear uncertain multi-agent systems subject to an uncertain leader system. In comparison with the case where the leader system is known exactly, we need to combine adaptive control and robust control techniques to deal with the unknown parameters in the leader system. By employing a distributed internal model for the multi-agent system, we first convert the problem into a global adaptive robust stabilization problem of a distributed augmented system. Then we further solve the global adaptive stabilization problem of the augmented system by a distributed adaptive control law. Finally, we show that the unknown parameters in the leader system can be exactly estimated provided that the distributed internal model is of minimal order. [Copyright &y& Elsevier]