Study On Combustion Control Of Hot Water Boiler Based On Neural Network PID

Combustion control of hot boiler has important meaning for heatingsystem regulation, boilers' safe and high-efficiency running and energy saving.Chain grate stroke hot water boiler, which is hard to control, is a nonlinearobject with bigger time lag, changing load and instability. How to control theboiler combustion system to ensure suitable quantity of heat for changing loadand guarantee boiler's safe, economical, and stable running is a hot-spotsubject in the area of energy and control.The production principle and characters of hot water boiler with chainstoker and water pump are analyzed. Combined with the requirements of heatproduction, living traditions and national regulations, a method according tothe air temperature to establish the temperature of supply water is put forward.The constrained model of supply-water temperature for boiler automaticcontrol is built.The characters and adjusting methods of boiler combustion are researchedand the parameters' relationship in combustion control system is established.The control blue print of combustion system composed of three independentsubsystems, which are fuel control system, input-wind control system andoutput-wind control system is established and the subsystems' systematicstructures are given.The control principle of neural network PID (proportion, integral anddifferential) algorithm is expounded. Its control characters are analyzed andcompared with traditional PID's through simulation. Based on the characters ofsubsystems, fitting algorithm is selected for each subsystem.According to the difficulties in fuel control, a new kind of neural networkPID algorithm with predictive model is put forward, which can be utilized innonlinear fuel control system with big time lag and excessive disturbances. Thesimulation research of the new algorithm in fuel control system is carried outand the flow chart and program codes are given. The experimental results showthat the new algorithm achieves satisfied effect in fuel control system.