Optimum Gain Selection of Sliding Mode Control using Grey Wolf Optimization Technique

Abstract

This paper presents a new methodology for the selection of controller gains in sliding mode control. The goal is to create an adaptive gain sliding mode control mechanism that is robust to uncertainty and perturbations without knowing the bounds of the uncertainties (only the boundedness feature is known). In addition, the approach should work with higher-order sliding mode controllers. The proposed method uses Grey Wolf Optimization (GWO), a new evolutionary algorithm that has been proved to outperform existing swarm intelligent optimization algorithms. Optimization characteristics assures that the gain is not overestimated. The effectiveness of the proposed approach is proven in an example using a robotic manipulator.