| In the electric power market, for the energy crisis has become increasingly prominentand the awareness of people to the importance of environment are raising, energy-savingand environmental protection is the eternal subject to coal-fired industry. At present,firepower plant still plays a powerful place in our country, about 70% electricity fromfirepower plant. The nation constitutes a series of reform measures and environmentalrequirements to thermoelectricity station that improve economic benefit and limit pollutantemissions to atmospheric. Modern power plants face to many pressures such as loweremissions, improve efficiency and reduce running cost.Therefore, domestic and foreign experts and scholars have being studied on boilercombustion system, and trying to solve the above problems. Aimed at complex propertiesof the boiler combustion system, for instance multi variables, multiple interferences,strong coupling, longtime delay etc, this paper makes an in-depth study and the supportvector machine (SVM) is introduced to the power plant boiler combustion optimizationmodeling. Using least square support vector machine (LSSVM) build the power plantboiler combustion optimization model and using Gravitational search algorithm (GSA)optimize model parameters to obtain the optimal model. Then using the GSA again findsout the best input-parameters combination of the optimal model in order to reduceemissions and improve the boiler efficiency. Finally, operator can adjust the relevantinput-parameters by the optimized parameter values, so that realize optimized operation ofthe boiler combustion system.Using above theories build the boiler combustion optimization model and carry outexperimentation simulation. The results show that the system of boiler combustionoptimization realizes the boiler high-efficiency and low-pollution emissions by integratedmodeling. This modeling method provides effective theoretical basis for building boilercombustion optimization system. |
PDF Full Text Request