Design of a hybrid adaptive fuzzy and neural network controller for nonlinear laser alloying process

Abstract

The laser surface alloying quality may vary significantly with respect to the process parameters variation. Hence, a feedback control system is required to monitor the operating parameters for yielding a good quality control. Since this multi-input and multi-output system has nonlinear coupling and time varying dynamic characteristics, it is very difficult to establish an accurate process model for designing model-based controller. Hence a hybrid model-free adaptive fuzzy and neural network controller (HAFNC), which combines an adaptive rule with fuzzy and neural network control, is employed in this study to overcome the difficulty. It has on-line learning ability for responding to the system's nonlinear and time-varying behaviors. Since this model-free controller has simple control structure and small number of control parameters, it is easy to implement. Two HAFNC controllers are designed for tuning the laser power and the traverse velocity simultaneously to tackle the absorptivity and geometrical variations of the work pieces. Simulation results based on semi-experimental system dynamic model and parameters show that the good surface lapping performance is achieved by using this intelligent control strategy.