Abstract: Abstract:Aiming at the problems existing in the 2×300MW unit natural circulation boiler of Mudanjiang Second Power Plant of Huadian Energy Company Limited, the operation optimization system was proposed. The system takes the test data as core, the influence of multiple variables on production indexes are considered synchronously in operation. A mass of operation data were collected by experiment. Using several mathematic tools, such as the artificial neural network, the statistic regression analysis and the fuzzy mathematics, a nonlinear mathematical model of the multiple variables and multiple objectives was established. The model can provide operating method and realize the online openloop guidance and closeloop control for boiler optimal operation. The application of the boiler optimal operation system in Mudanjiang Second Power Plant shows that the net coal consumption rate and the nitric oxide emission of the power plant were reduced under multiple limiting conditions, and the economic efficiency of the boiler combustion was improved

Key words: Keywords:boiler；operation optimization system, combustion adjustment, nitric oxide (NOx), multiobjective optimization, model predictive control