基于NSGA-II的水轮机固定导叶多学科优化设计

Multidisciplinary design optimization of stay vane in hydraulic turbine based on NSGA-II algorithm

水力性能和结构强度是固定导叶设计时的两大重要因素。本文提出了一种优化设计方法,在同时考虑水力性能和结构强度的基础上对固定导叶外形进行了优化设计。该方法以固定导叶的骨线形状、相对厚度分布规律和最大厚度作为优化变量,以固定导叶通道的总压损失、导叶上的最低静压值以及导叶在水流作用下的最大静应力值作为优化目标对固定导叶进行了优化设计。优化算法采用NSGA-II多目标遗传算法,该算法能有效的处理不同目标函数之间的冲突关系,从而找到使各个目标函数均有提高的最优解。应用本方法对某电站的模型水轮机固定导叶进行了优化设计,优化后的导叶通道总压损失减少了39.8%,导叶上的最低静压值上升了4.5%,最大静应力值下降了6%,这表明该方法具有一定的工程应用价值。

Hydrodynamic performance and structural strength are two key factors in design of stay vane. This paper develops an optimal method that improves the hydraulic performance and structural strength of stay vane simultaneously. This method uses mean line shape, relative thickness distribution and maximum vane thickness as its optimization variables, and its objective functions for optimization of stay vane geometry are the total pressure loss of flow passage, the lowest static pressure and the peak stress of stay vane. A NSGA-II algorithm is adopted for optimization calculation. It is effective in solving the conflict between different objective functions, and it can find the optimal solution and improve all objective functions. Application of the developed method to stay vane design for the model turbine of a hydropower station, shows a reduction 39.8% in the total pressure loss of flow passage, a rising 4.5% in the lowest pressure on the vanes, and a reduction 6% in the peak static stress of the vanes. Thus, the optimal method is a practically useful method for engineering design, optimization algorithm