An adaptive fuzzy neural controller

The former adaptive neural controller either supposes the plant is linear and has identity function([1]), or chooses the identification transfer function mechanism to replace the real plant([2]) in order to use the BP algorithm. These methods are only about how to use BP algorithm in control field, and the shortcoming of BP algorithm itself still existed. We present an adaptive fuzzy neural controller which combines fuzzy logic and BP algorithm, highlights their advantages but avoids their disadvantages. The presented controller([1]) can control the situations that the former controller can not control. The simulation results of third order plant with disturbances and dead times show the validity of the presented controller.