Model of Adaptive System of Neuro-Fuzzy Inference Based on PID- and PID-Fuzzy-Controllers

Abstract

The goal of this work is to develop a model of adaptive system of neuro-fuzzy inference based on PID and PID-FUZZY-controllers, which would allow connecting formalized and informal knowledge in the design of modern automated and automatic control systems for technical processes. To achieve this, we developed an approach to the control of a technical object using an adaptive system of neuro-fuzzy inference. The main control elements of the developed adaptive system of neuro-fuzzy inference are the PID and PID-FUZZY-controllers, as well as the classical and fuzzy control models designed on their basis. The interaction of the two models is provided by the developed hybrid control system. Resulting from the interaction of the two models, the rules base of the fuzzy controller is automatically formed, based on the knowledge about the object obtained when it is controlled by the classical controller. This completely excludes the expert from setting and tuning the parameters of the fuzzy controller. In the developed adaptive system of neuro-fuzzy output, the deviation signals and the deviation and control differential in the classical model are used as data for building a hybrid network. Deviation and control signals in the fuzzy model with automatically generated fuzzy inference rules are used as data to test the hybrid network in order to detect the fact of its overtraining. Thus, the application of the adaptive neuro-fuzzy inference model based on PID and PID-FUZZY controllers allows the effective control of a technical object in the conditions of uncertainty #CSOC1120.