基于多目标进化算法的差动式调压室优化研究  

Multi-objective evolutionary optimization of large differential surge chamber

作　　者：缪明非[1] 张永良[1] 马吉明[1] 江春波[1] 

机构地区：[1]清华大学水利水电工程系水沙科学与水利水电工程国家重点实验室,北京100084

出　　处：《水力发电学报》2010年第1期57-61,185,共6页Journal of Hydroelectric Engineering

基　　金：国家自然科学基金(50539070);973项目(2006CB403304)

摘　　要：本文将多目标进化算法引入到大型差动式调压室的优化中。编制了双升管差动式调压室涌浪数值计算的程序,选择分别反映结构安全风险和水力性能的最不利组合工况下的升管主室最大正压差和调压室最低涌浪水位作为优化目标,采用非支配排序遗传算法(NSGA-Ⅱ),将NSGA-Ⅱ与水力过渡数值计算结合起来,对差动式调压室的升管阻抗孔口面积、回流孔口面积、升管面积,升管溢流堰高程进行优化,并应用于锦屏二级上游差动式调压室的体型优化中,得到了升管主室最大水位差与调压室最低水位之间的Pareto前沿。Pareto前沿上部分方案比现有方案在水力性能和结构安全风险这两个目标上略为优越。The paper studies the evolutionary multi-objective optimization of the large differential surge chamber. The lowest water level in surge tank and the maximal pressure head between the riser and main tank are the two optimization objectives which represent hydraulic performance and structural risk respectively. Based on analysis of methods of multi-objective optimization, the non-dominated sorting genetic algorithm （NSGA-Ⅱ ） is applied to optimize the design parameters including the height of overflow weir of riser, the area of riser, the area of throttled orifice at the bottom of the riser, the area of interconnected orifice between the riser and main tank. The Pareto front of the two objectives are obtained, tradeoff of the two objectives are performed. NSGA- Ⅱ is applied to the optimization of Jinping-Ⅱ hydropower station. The Pareto front of the two objectives is obtained. Some schemes in the Pareto front are better than the existing scheme in the two objectives.

关 键 词：差动式调压室 多目标进化算法 升管 结构安全风险 

分 类 号：TQ028.8[化学工程]