The Research On The Combustion Process Optimizztion Of The Circulating Fluidized Bed Boiler Based On The Artificial Bee Colony Algorithm

With the increase of the demand for electricity, and the implementation of thenational policy for energy conservation and emission reduction, power plants are forced toimprove thermal efficiency and to reduce the emission of pollutants. The optimizationtechnology for plant boiler combustion not only can improve the thermal efficiency butalso can observably reduce the emission of pollutant, so it turns to be a significant researchsubject.In this paper, aiming at a domestic300MW Circulating Fluidized Bed Boiler, amodified artificial bee colony algorithm is put forward to do a deep research in combutionoptimization and then applied in the following two ways:1) The comprehensive model ofthe NOx emission and the thermal efficiency is built with the combined least squaresupport vector machine, in which the operation data is the input;2) The boiler’scontrollable parameters are optimized to get a best matching of variables. When workingin this condition, the NOx emission is notably reduced and the thermal efficiency isincreased. What’s more, compared with ABC algorithm and PSO algorithm, the modelbuilt in this article shows a better prediction of the NOx emission and provide a goodrecognition capability and generalization ability as well. Besides, the high speed of themodified artificial bee colony algorithm and its advantage in optimizing databaseperformance are well reflected simultaneously.Through the simulation test, the optimized NOx emission concentration has beeneffectively reduced, and the boiler thermal efficiency has been improved accordingly. Theleast squares support vector machine (LSSVM) model based on improved artificial colonyalgorithm has high precision and generalization ability, which shows the engineeringpractical value.