A distributed hybrid sliding mode control system for a class of nonlinear mechanical systems

This paper presents distributed hybrid sliding mode control (SMC) technique for a class of nonlinear mechanical systems. The proposed technique reduces observer-controller gains from classical output feedback SMC approach via reducing the level of parametric uncertainty. To this end, we distribute uniformly the compact set of unknown parameters into a finite number of smaller compact subsets. Then we construct a family of distributed candidate SMC for these smaller compact subsets. The Lyapunov-like energy function is employed to identify an appropriate candidate SMC that closely estimates the plant at each instant of time. The key idea of the proposed design is to allow classical SMC to be switched into a distributed candidate SMC that best approximates the plant among a finite set of the distributed candidate SMC. The method is evaluated on a 2-DOF robot manipulator to demonstrate the effectiveness of the theoretical development.