A New Strategy for Batch Reactor's Temperature Control

Model based batch reactor's temperature control algorithms for special plants or systems have been used widely. Adaptive control is also used to handle the model changing problems. This paper presents a new model free adaptive control that is based on expert system, fuzzy control and neural networks. Its main structure is a model free neural networks and the initial values of the neural networks parameters are given by an expert system. The learning rate and other adaptive parameters are updated by the fuzzy control technology.