An Application Research On Optimization Of Power Plant Coal-fired Boilers Combustion By Free Search Algorithm

Energy is the material basis for the existence and development of human society.Along with the development of the national economy, the demand for electricity increasessharply. In recent years, people’s awareness of environmental protection continuouslyimprove, and PM2.5is included in the pollutant reduction evaluation system. In early2013, persistent fog and haze change the thermal power plants to the target. Therefore,thermal power enterprises are facing to not only reduce operating costs, but also lessenemissions of pollutants.In recent years, scholars have conducted extensive research about the combustionoptimization technology at the theoretical level and software technologies, and haveachieved certain achievements. The running boilers of the power plant mostly arepulverized coal boilers whose combustion system’s characteristics are multi-variable,non-linear, strong coupling, large delay characteristics. Thus, this paper collected the rundata of a300MW pulverized coal boiler and the burning coal quality data. The collecteddata is preprocessed. On this basis, Free Search algorithm (FS) and the cross-validationmethod are applied to the optimization process of super-parameter of the Support VectorMachine (SVM) model. Two models are established for the output of NOx emissions andcombustion thermal efficiency. Based on the two models, FS algorithm is used in theoptimization of adjustable parameters, and the targets of the combustion are to reduceNOXemissions and improve the efficiency of the heat of combustion. The effect is verygood. Finally, the NOx emissions and combustion thermal efficiency are integrated as onetarget, and the FS algorithm is used in the optimization of adjustable parameters to achievea multi-objective optimization of the boiler combustion system. By adjusting the ratio ofthe weight, the optimization effect can meet the running demand of different powerstations.The study is used in the simulation experiments. The simulation results show that theestablished models have high precision, generalization ability. And the effect of theoptimization is very obvious. Moreover, the goal of reducing NOx emissions and improving the thermal efficiency of combustion optimization is achieved.