Application Of Fuzzy Neural Network Control To The Main Steam Temperature Control System In Coal-fired Power Stations

Boiler main steam temperature is one of the main parameters of the whole system duringthe operation of the boiler in heat-engine plants, and it has great influence to the safety andeconomic operation of the power plant. The primary regulations of boiler superheatedtemperature is steam side regulation and the flue gas side regulation, but because the way ofadjusting the flame center position to the flue gas side steam temperature regulation is notmeticulous, so it is often used in conjunction with the Spray desuperheating. Regulation of thesteam side temperature mostly is desuperheater spraying water to the superheated steam. Atthe present stage, heat-engine plants popularly use the conventional PID cascade controller.But the development to large capacity, high parameter, high efficiency of boiler units ofmodern power plants, and characteristics of large inertia, large delay, nonlinear andtime-varying of the steam temperature object, which make the traditional PID cascade controlis difficult to achieve the optimal control under various engineering conditions, and cannotsatisfied the demand of modern industrial production. So the new control is urgently neededto effectively controlling of the main steam temperature of boiler at this stage.Neural network has the powerful adaptive learning ability, which can make the net adaptwell to changes of controlling environment, and select feature automatically by simple toolsto produce useful data, but the mean of expression of the relationship between input andoutput of neural network is not easy to be accepted; On the contrary, the fuzzy control has noability of learning and adapting, but it can get expert knowledge expressed by language easily,which can make great real-time control to the system by the experience and knowledge ofexpert rather than the accurate model.In this paper, according to the current situation of power plants in China and thedevelopment of intelligent control, I apply the fuzzy neural networks to the main steamtemperature of boiler, with the power plant boiler main steam temperature as the researchobject, and solved the problems and difficulties in the traditional PID control, with achievingthe effective control, raising the control level of steam temperature of boiler in thermal power plant. Finally, the simulations results show that PID control based on the fuzzy neural networkhave a good self-tuning effect, so it has a certain application prospect. Go a step further，basedon the ANFIS,I emulated a new self-tuning fuzzy neural networks to the steam temperature ofboiler in thermal power plant, getting rid of the dependence to the expertise, and through thesimulation, proving the feasibility of this new system.