Genetic algorithm based nonlinear self-tuning fuzzy control for time-varying sinusoidal vibration of a magnetorheological elastomer vibration isolation system

Abstract

Protecting high-tech facilities from micro-vibrations in the working environment has demanded intensive research in recent years. One of the most promising devices proposed for facilities protection is the magnetorheological elastomer (MRE) isolator. To explore fully their potentials in the real-time control implementations, control strategy plays an important role. This paper proposes a nonlinear self-tuning fuzzy controller (STFC) for a MRE micro-vibration isolation system exposed to time-varying (variable frequency and amplitude) sinusoidal excitations. The silicone rubber matrix MRE isolator with low initial modulus is employed to suppress the time-varying vibration at low frequency, the performance of the proposed isolator is evaluated by sweep frequency experiments. Different from the conventional fuzzy controller (FC) with fixed parameters, which is only effective for the vibration with a certain frequency and amplitude, a STFC composed of semi-active FC and self-tuning law can suppress time-varying vibration with the wide frequency and amplitude range. In order to achieve the adaptability of the proposed controller to time-varying excitation in real time control, the nonlinear self-tuning law is designed by using genetic algorithm to optimize scaling factors. Both numerical simulations and experiments are conducted to verify the effectiveness of the designed controller. Meanwhile, the control effects between STFC and conventional FC are compared in acceleration attenuation to demonstrate the superiority of the proposed STFC. The results show that the designed nonlinear STFC is more effective for suppressing the time-varying vibration with the amplitude of 0.5–3 m s−2 and frequency of 45–60 Hz.