作　　者： ; ; ; ;

机构地区： 清华大学机械工程学院热能工程系

出　　处： 《动力工程》 2005年第5期676-679,706,共5页

摘　　要： 根据某200MW单元机组燃煤锅炉热态试验数据,采用人工神经网络建立了电站锅炉运行效率和污染排放的预测模型,并根据电站运行利润最大化和经济环保相协调的要求,提出了3种锅炉运行优化目标。在上述模型和优化目标的基础上,采用遗传算法对试验机组进行2min1次的寻优,实现了问题的实时求解。以上述算法为核心,通过与锅炉机组DCS系统进行双向通信,构成了锅炉运行优化监督控制系统,实现了电站锅炉环保与经济相协调的实时优化控制。 Based on hot state test data of a certain coal fired 200MW block unit, predictive models for both boiler operation efficiency and pollutant emission have been developed with the help of artificial neural networks. According to the requirement of maximized profitability of power plant operation, optimization targets for harmonized boiler operation effieiencv and environmental protection under three different conditions have been advanced. On the basis of these models and with the optimization target mentioned, and by using a genetic algorithm for optimizing every 2 minutes, the real time solution of the problem is found. With this algorithm at its core, bv establishing a mutual interconnection with the boiler＇s DCS system, optimized monitoring control of boiler operation is attained, and with it, real time optimized control of both boiler efficiency and environmental protection. Herewith the unit＇ s maximal profitability can be warranted. Figs 2, tables 3 and refs 9.

关 键 词： 动力机械工程 协调优化 锅炉效率 污染排放

领　　域： [动力工程及工程热物理]