The Research Of Detection Method Of Boiler Combustion State Parameter Based On Information Fusion

The problem of energy conservation, consumption reduction as well as exhaust reduction is essential for coal burning units, it has emerged as a key issue, for heat engine units, that how to improve the efficiency of electric power generation by the action of performance optimization. The combustion state of boilers may prove economically and safely attractive to quite a wide range of units and thereby determines the optimization. Furthermore, state parameter detection is the primary basis for the optimizing control action of boiler combustion.Currently, on-site detection information correlative to combustion state is insufficient relatively. However, information fusion and soft sensor have offered a new solution to figure it out. The application of SIS (Supervisory Information System) in power plants gave us abundant data information and inspiring verification platform. Based on the accurate mixture ratio between air and fuel, the main work of this dissertation can be presented as follows:1.To guarantee the accuracy and reliability of air and fuel flow measurement, a new method was developed based on heterogeneous data fusion. Redundant information, consist of relevant heterogeneous data, was employed to improve the accuracy and reliability of measurement while the heterogeneous data fusion means combining statistical modeling to data fusion was also adopted. The effectiveness of the means was verified through actual data calculation.2. Soft sensor signal of oxygen content in boiler flue gases based on information fusion was constructed. At the same period, via mechanism and statistical analysis, the soft sensor model for oxygen content was established with air leakage, coal quality, air flow and fuel flow coming into consideration. Similarly, the accuracy of this measurement was verified by actual data calculation. As a typical example combining modeling by mechanism to information fusion, it has come into its own on SIS and DCS.3.Aiming at variable coal quality in power station, on-line soft sensor was developed. The variety of coal quality will impact air fuel ratio. The analytic expressions for net calorific value, moisture content and grindability of coal were established by both energy and mass balance laws. Based on comparisons of actual calculated results and statistical data in analysis reports on coal quality of power plants, the rationality of results were proved.4.Aiming at frequent fluctuation of unit load command and inaccurate mixture ratio during dynamic process between air and fuel, dynamic performance of mill pulverizer system was researched as well as air supply system. The mill dynamic model was established and its online modification for fuel command gain and inertia was performed with coal-fired calorific value and grindability index. Simultaneously, non-linear dynamic model was also built for auxiliary air system by the means of mechanism and statistical analysis. In addition, in order to solve the problem between air flow response speed and air pressure response fluctuation, coordinated control scheme of auxiliary air system was designed while actual data simulation ensures the accuracy of model and the feasibility of control.