Self-adaptive interval type-2 neural fuzzy network control for PMLSM drives

This paper proposes a self-adaptive interval type-2 neural fuzzy network (SAIT2NFN) control system for the high-precision motion control of permanent magnet linear synchronous motor (PMLSM) drives. The antecedent parts in the SAIT2NFN use interval type-2 fuzzy sets to handle uncertainties in PMLSM drives, including payload variation, external disturbance, and sense noise. The SAIT2NFN is firstly trained to model the inverse dynamics of PMLSM through concurrent structure and parameter learning. The fuzzy rules in the SAIT2NFN can be generated automatically by using online clustering algorithm to obtain a suitable-sized network structure, and a back propagation is proposed to adjust all network parameters. Then, a robust SAIT2NFN inverse control system that consists of the SAIT2NFN and an error-feedback controller is proposed to control the PMLSM drive in a changing environment. Moreover, the Kalman filtering algorithm with a dead zone is derived using Lyapunov stability theorem for online fine-tuning all network parameters to guarantee the convergence of the SAIT2NFN. Experimental results show that the proposed SAIT2NFN control system achieves the best tracking performance in comparison with type-1 NFN control systems.