Robust H-infinity controller synthesis using multi-objectives differential evolution algorithm (MODE) for two-mass-spring system

Abstract

One of the major problems associated with H-infinity controller design is the challenges of weighting function parameters' tuning. For a given control problem, it is requires to select appropriate weighting function parameters to satisfy both the stability requirement and time domain specifications. The conflicting nature of these objectives has made such design exercise a tedious endeavor. In this study, the optimal-tuning of the H-infinity controller parameters using multi-objective differential evolution (MODE) technique is proposed, and apply for control of a two-mass-spring system which represents a typical benchmark robust control problem. The performance of the resulting robust controller is evaluated both on the nominal plant and perturbed plants with parameter variation. The simulation results indicate a good compromise between the robustness and time domain performance.