| Popis: |
Experimental and simulation investigations of combustion optimization were performed on a 125 MW tangentially, anthracite- fired boiler. The experiments including forty cases by burning three pure and five blended coals were firstly carried out under the different O2 concentrations in flue gas and distribution modes for secondary air. The results show that a reverse V-shaped operating for the secondary air mode and O2 concentration of 3%–4% is favorable to decrease the unburned carbon content in the fly ash and obtain the high boiler efficiency. After that, the BP artificial neural network technology was adopted to develop a simple network model between coal quality, boiler operating conditions and the boiler efficiency. The most important and representative parameters such as V ar and Q net, ar from approximate analysis, distribution mode for the secondary air and O2 concentration in flue gas at the furnace exit, were chosen as input parameters. The distribution mode for secondary air was denoted by the ratio of the damper opening values between upper and lower secondary air. Then a preprocessor method of approximate optimizing, based on the above model, was proposed to optimize the combustion performance of the boiler and establish an information library of combustion optimization for the boiler. Finally, validation tests were conducted to check the applications of the optimum information library. |
