基于试验遗传算法的水电站加劲压力埋管优化设计

水电站地下埋管的结构最优化设计是一个多维混合变量的非线性优化问题,采用常规优化方法求解有较大困难,为此提出基于试验优化设计思想的改进遗传算法———试验遗传算法。实例表明,由于该算法能自动调整计算精度,更好地保持种群多样性而易获得全局最优点,且计算简单、高效,对水利水电工程中常见的高维非线性优化问题适用性强,具有较高的推广应用价值。

Optimization design for embedded stiffened penstock of hydropower station base on experimental genetic algorithm

It is a multi-variables-mixed non-linear optimization problem in designing the structure of the embedded stiffened penstock of hydropower station, and it is difficult to be solved with general optimization methods. Therefore, a new improved genetic algorithms based on the concept of experimental optimization method - experimental genetic algorithm is put forward herein. The actual case shows that the improved algorithm has the features of better optimizing efficiency and good adaptivity of the calculation precision. It also has both the ability to keep the individual variety better to get the global optimization, and good applicability to the high dimension nonlinear optimization in the area of water resources and hydropower engineering, and then is valuable to be introduced into the engineering field concerned.