Dynamic neuro-fuzzy control of the nonlinear process

This research combines the dynamic neural networks with the fuzzy associative memory (FAM) to find a better model for nonlinear control problems. The proposed model consists of three major parts: the action networks, the critic networks and the fuzzy membership adjustment procedure. The dynamic neural networks are used as the main controller of the system. The FAM determines the performance of the main controller and sends the correction signal back to the neural networks. The function of the fuzzy membership adjustment procedure is to improve the quality of the FAM output. The proposed model is tested on real-life processes, and satisfactory results are obtained.