Adaptive backstepping control of linear induction motors using artificial neural network for load estimation

Abstract

Linear induction motors (LIMs) make performing a direct linear motion possible without any mechanical rotary to linear motion transforming parts. Obtaining a precise mathematical model of such type of motors presents a difficulty due to time varying parameters and external load disturbance. This paper proposes an adaptive backstepping controller structure based on lyapunov stability for controlling a LIM position. Which can guarantee the annulment of position tracking error, despite of parameter uncertainties. Parameter update laws are extracted to estimate mover mass, friction coefficient and load force disturbance, which are assumed to be constant parameters; as a result, compensating their undesirable effect on control design. Then, load disturbance estimate is replaced with an artificial neural network (ANN) to reduce the estimation error. The numerical validation has shown better performance compared to the conventional backstepping controller, and proved the robustness of the proposed adaptive controller design against parameter changes.