Direct and Indirect Adaptive Inverse Control Design for Non-Minimum Phase EHA System with Periodic Disturbance via FASS-NLMS Algorithm
Keywords: Adaptive Filter, Adaptive Inverse Control, Fuzzy Systems, NLMS, Stochastic Gradient Descent
AbstractIn this paper, is performed the comparative analysis of performance between the Direct and Indirect Adaptive Inverse Control applied to a non-minimum phase Electro-Hydraulic (EHA) system in the presence of a periodic disturbance signal. The performance of an adaptive inverse controller is influenced by the trade-off between the convergence speed and the steady-state MSE during the update of the estimate of the weights vector. Aiming to propose a new optimization algorithm based on stochastic gradient descent, in this paper, a new version of the NLMS algorithm with adaptive step size is proposed, with the goal of obtaining a good trade-off between convergence speed and the steady-state MSE. For this, the step size is adapted by a Mamdani Fuzzy Inference System in function of the squared error and of the instant of normalized time by the Min-Max method. Computational results illustrate the efficiency of the proposed optimization algorithm in the design of these two approaches of adaptive inverse control.