Adaptive type fuzzy neural-network (FNN) backstepping motion control strategy based on sliding-mode scheme for induction motor drives with robust position tracking capability

In this paper, an adaptive type fuzzy neural-network (FNN) backstepping controller is proposed for the position tracking of a mechanical system driven by an induction motor. The primary concept for integration of the control strategies is described and derived. The demonstration mechanical system is set up as a single link structure fixed on the shaft of the induction motor to form a variant load system based on robotic application. The backstepping methodology provides a simple design procedure for an adaptive control scheme and can be integrated into the approach for defining the sliding surface based on the sliding-mode control related to the robustness criterions. Hence, the backstepping control can be extended further to be an adaptive sliding-mode controller easily. A four-layer FNN is also applied for enhancement of the perfect lumping of uncertainties. The verification results show that position control performance of this induction motor drive is stable and robust under parameter variations and external disturbances by FNN learning capability as well as adaptive features from the overall control scheme.