Advance application of fuzzy Neuro modeling and simulation methods using NEFCON model for throttle valve governing in turbine in power plant

Abstract

This paper presents application of advance fuzzy-neural modeling and simulation of throttle valve control problem in turbine and the optimization of fuzzy rules previously developed using NEFCON model2. The conventional method of throttle control is non-automatic and a automatic type PID controllers3. These controllers are based on mathematical models in which the control system is described using one or more differential equations. These systems are products of decades of development and theoretical analysis. In many cases the mathematical model of the control process either not exist or is very complex or may be too expensive. The hardware and software of fuzzy systems are cheap and such systems can be easily upgraded by adding new rules. The group of researchers presented different types of architectures of Neuro fuzzy and fuzzy neural networks. Advance fuzzy modeling environment is presented for system design. Development of fuzzy sets and fuzzy rules for throttle valve control need rule base design incorporating the human intelligence. To obtain the optimized output from the previously designed system rule base a combination of fuzzy and neural network is proposed. Looking into the optimizing property of NEFCON network the fuzzy system is combined with it and a fuzzy neural system is generated6-7. The simulation results of proposed fuzzy-neural system used for throttle valve control is presented and discussed2.