Design and analysis of novel Chebyshev neural adaptive backstepping controller for boost converter fed PMDC motor

Abstract

An adaptive backstepping Chebyshev neural network controller (ABCNNC) is proposed for the boost converter fed PMDC motor to track the angular velocity. The computational complexity of the neural network is avoided by the use of Chebyshev polynomials as the basis function. The online weight update of the Chebyshev neural network (CNN) is designed for the closed loop system based on the Lyapunov stability analysis to obtain the asymptotically stable system. A detailed analysis of the steady state and transient performance is performed and results are compared with that of conventional PI controller and radial basis function neural network controller (RBFNNC). To ensure the robustness of the proposed ABCNNC, it is being analysed for a wide range of variations in load torque and the set point changes and it is validated by comparing with the conventional PI control approach and RBFNNC. Comparison of results validates that the proposed ABCNNC shows the enhanced transient and steady state responses for the uncertainties caused by disturbances, than conventional PI controller and RBFNNC.