Improved Adaptive Hybrid Compensation for Compound Faults of Non-Gaussian Stochastic Systems

Abstract

This study investigates an adaptive estimation and hybrid parallel distribution compensation for non-Gaussian stochastic systems with actuator-sensor compound faults. Type-II fuzzy theory approximates nonlinear dynamical systems with non-Gaussian stochastic outputs and inconsistent response under harsh environments. An estimation observer using fuzzy and improved adaptive laws is proposed to accurately estimate different amplitudes of an actuator fault. In a fault-tolerant controller, the indirect passive compensation factors shield the sensor from the initial fault; therefore, the controller actively compensates for the actuator fault using the estimated information. The output probability density functions then match the expected values. Finally, the Lyapunov functions prove the robust stability and the simulation results indicate the effectiveness of these hybrid methods.