本文選題：火電廠 + 負荷優(yōu)化分配��； 參考：《熱力發(fā)電》2014年12期

【摘要】：在傳統(tǒng)的火電廠經(jīng)濟負荷分配模型基礎上,綜合考慮全廠供電煤耗率、污染物排放量以及全廠負荷升、降時間3個目標,構建了廠級負荷優(yōu)化分配的多目標模型。將差分粒子群混合算法發(fā)展為一種新型的多目標進化(MDPA)算法,即利用擂臺賽法和凝聚層次聚類分析方法分別構造和修剪非支配集,同時加入精英保留策略,保留進化過程中的極值點。將該算法應用于以經(jīng)濟、環(huán)保、快速3個目標為多目標的廠級負荷優(yōu)化分配,并與基于非支配排序的多目標優(yōu)化(NSGA-Ⅱ)算法進行對比。結(jié)果表明,MDPA算法較NSGA-Ⅱ算法收斂速度更快,解集分布更均勻。
[Abstract]:On the basis of the traditional economic load distribution model of thermal power plant, a multi-objective model for optimal load distribution of power plant was constructed by considering three objectives: coal consumption rate of power supply, pollutant emission and load increase and drop time of the whole plant. The differential particle swarm optimization (DPSO) hybrid algorithm is developed into a new multi-objective evolutionary MDPA algorithm, in which the non-dominated sets are constructed and pruned by using the beating table method and the condensed hierarchical cluster analysis method, and the elite retention strategy is added. Preserve the extremum of evolution. The proposed algorithm is applied to plant load optimal allocation with three objectives of economy, environmental protection and speed, and is compared with the NSGA- 鈪，

本文編號：1809870