Adaptive hybrid neural fuzzy controller using augmented error method

Abstract

The design of fuzzy controller can be supported by comparing the signal with the neural network controller. Such approaches are usually called hybrid neural — fuzzy controller or multi controller. The hybrid model is able to compare the signal from the fuzzy controller and neural network learning with back propagation method. In this paper the plant is a DC motor base assembly with Pittman gear head servomotor using in robot and another applications, in order to evaluate the system performance when the motor load is changing. Due to this change the speed of the motor will be decreasing and the plant parameter is changed. Therefore, adaptive hybrid neural fuzzy controller is designed to adapt this system change using augmented error method.