Experimental fuzzy control improvement with surface response methodology for power electronics

Surface response methodology is used in this paper to improve the dynamic performance of pre-established settings for PID-like fuzzy controller dedicated to under-damped behaviour systems with hard load variations. Only two open loop successive step input tests are required for controller parameter tuning with only one voltage sensor. Using information given by experimental designs, the experimental response surface is estimated. As a consequence, the analytical optimum could be found between the chosen levels of the experimental parameters. Dynamic performance on the output voltage are significantly improved but the robustness versus the reference level decreases