Parameter Prediction And Combustion Control For Region Heating Station Based On Fuzzy Theory

Based on the theory of fuzzy neural network and fuzzy intelligent control and according to the analysis of practical operation for region heating station, the paper presents a new control method of 'Double Fuzzy' control system for this complicated nonlinear system: parameter prediction by fuzzy neural network and intelligent control by fuzzy controller. Chun Jiang region heating station at QingDao is regarded as study object. This key technology to control region heating station is performed to study with good effect in terms of field experiment and computer simulation.According to the analysis to the structure and character of region heating station, two subsystems are presented: combustion capacity-exchanged subsystem and heat-net burdened subsystem. Based on this double subsystem of boiler, the collected multi-parameter of weather and practical operational data are used to build sample data, modified fuzzy neural network (FNN) and neural network-driven fuzzy reasoning (NNFR) are built to predict water temperature of exportation and inlet-the key parameters of heat-net burdened subsystem and to repredict the operational parameters of combustion capacity-exchanged subsystem with upper predicted results. The ultimate predicted results are corresponding to practical operational data of boiler and the feasibility of this method is proved. The first step of 'Double Fuzzy' control system is implemented.Toward the second step of'Double Fuzzy' control system-intelligent control by fuzzy controller, the paper decomposes the combustion system of boiler to three subsystem: give-coal subsystem, blow-air subsystem, tow-air subsystem. According to the practical operational data and specialized experience fuzzy control rules are required. Every subsystem's fuzzy controller is designed and simulation is also performed to study the controllers' property by software. 'Double Fuzzy' control system for region heating station is implemented.To sum up the method of 'Double Fuzzy' control system presented for region heating station is proved to have good precision and practicable value according to the data comparison. Especially this method solves the model building with multi-parameter of weather including the direction of wind, the magnitude of windand the magnitude of cloud by fuzzy neural network in parameter prediction. Comparison to the static model of heat-net, this method has obvious advantage. Believably through more experiments in more region heating stations this method will have a widely prospect.