Hybrid RBF neural network–based sliding mode control for a magnetorheological elastomer isolator under variable frequency and amplitude excitations

Magnetorheological elastomers have attracted significant attention for semi-active vibration isolation applications due to their ability to rapidly change mechanical properties under an applied magnetic field. However, the nonlinear and hysteretic behavior of Magnetorheological elastomer materials poses challenges for achieving robust vibration suppression under varying excitation conditions. This study presents a hybrid control strategy that integrates a radial basis function neural network with a Sliding Mode Controller to enhance the vibration isolation performance of an MRE-based semi-active isolator. A MRE isolator with orthogonal magnetic flux configuration is designed and fabricated, and its dynamic characteristics are evaluated under harmonic excitation in the frequency range of 15–80 Hz. The proposed hybrid controller adaptively estimates system nonlinearities using the neural network while maintaining robustness through sliding mode control. Experimental and simulation results demonstrate that the proposed controller provides effective vibration attenuation while maintaining stable adaptive performance under varying excitation frequency and amplitude. The results show that the semi-active MRE isolator achieves a maximum vibration reduction of 55.8% near the resonance region under the investigated operating conditions, while maintaining stable performance across varying excitation frequencies. The proposed control framework provides an effective solution for handling nonlinear dynamics, and it offers greater adaptability in smart vibration isolation systems based on magnetorheological elastomers.