Apply Of Fuzzy Neural Net Control In Automatic Control Of Drug Adding System In Power Plants

Water and steam are the working media of thermodynamic facility in power plant .As there are impurities in the working media, it will erode the thermodynamic facility and threaten boiler safety. For maintaining the system's stable working, adding hydrazine or ammonia to feed-water and phosphate to boiler water are the common methods of adjusting water quality. While adding these drugs into the system, there also result in many matters So the amount of adding drugs must be controlled strictly, which affects the safe, economical run of power plant, especially the high parameter and large capability boilers are putted into production.Now, in our country manual mode is primary in dosing control system, which has not satisfied the demand of system work property and condition changing, and has aroused the fluctuation of medicament concentration in a quite wide range. For this reason, it is necessary to design and develop a dosing automatic control system that can adjust the dose automatically and continuously base with the change of water quality and avoid to erode the thermodynamic facility.In the autocontrol of dosing control system, the primary is the control tactic. Now, the main methods of dosing control are some single control arithmetics base on PID control. It's able to meet the demand of control when the water quality runs stably. While the load has severe fluctuation especially during the loading off, these single control arithmetics cannot meet and demand and result in bad control quality. Therefore base on the outstanding characteristic of time-varying and large time-delay of the system controlled object. Some control methods such as smith ,dahlin,IMC are applied in some fields. But because of some bugs, they cannot be applied more widely. Base on these, a new control method called "self-organized fuzzy control base on dynamic structure neural network "is proposed in this article. It syncretizes the strongpoint of Fuzzy control and Neural Net control, and attain optimization control with the adjusting the number of the net's layer. The MATLAB simulating research shows that this control method has satisfying control effect.