Adaptive Recurrent Functional-Link-Based Petri Fuzzy-Neural- Network Controller for a DSP-Based Induction Motor Servo Drive System

In this paper, an intelligent adaptive control system (IACS) for induction motor (IM) servo drive to achieve high dynamic performance is proposed. The proposed IACS comprises a recurrent functional-linkbased Petri fuzzy-neural-network (RFLPFNN) controller and a robust controller so that the developed adaptive control scheme has more robustness against parameters uncertainties and approximation errors. The RFLPFNN controller is used as the main tracking controller to mimic an ideal control law while the robust controller is proposed to compensate the difference between the ideal control law and the RFLPFNN controller. The proposed RFLPFNN model uses a functional-link neural network to the consequent part of the fuzzy rules. Thus, the consequent part of the proposed RFLPFNN model is a nonlinear combination of input variables. Moreover, the online structure and parameter-learning of the RFLPFNN are performed concurrently. The structure learning is based on the partition of input space and the parameter learning is derived based on the Lyapunov stability analysis and the back propagation method to guarantee the asymptotic stability of the IACS for the IM servo drive. In addition, to relax the requirement for the bound of minimum approximation error and Taylor higherorder terms, an adaptive control law is utilized to estimate the mentioned bounds. A computer simulation is developed and an experimental system is established to validate the effectiveness of the proposed IACS. All control algorithms are implemented in a TMS320C31 DSP-based control computer. The simulation and experimental results confirm that the IACS grants robust performance and precise response regardless of load disturbances and IM parameters uncertainties. Furthermore, the superiority of the proposed IACS is indicated in comparison with the Petri fuzzy-neural-network control system and traditional PID controller. Key-Words: Functional-link neural-networks (FLNNs), intelligent control, indirect field-orientation control (IFOC), induction motor, Lyapunov satiability theorem, Petri net (PN), fuzzy-neural-network, robust control.