基于改进粒子群算法的水泵水轮机多目标优化

Multi-objective optimization of pump turbine based on improved partical swarm optimization algorithm

为了提高带有长短叶片结构的水泵水轮机机组在运行中泵工况的运行效率,先对粒子群优化(PSO)算法进行线性变化的惯性权重及变化学习因子的改进,然后用改进PSO算法对水泵水轮机转轮结构进行优化.采取试验和数值模拟相结合的方式,以提高水泵水轮机泵工况的效率与扬程为目标,用近似模型和改进PSO算法结合的方法对转轮9个结构参数进行全局寻优.研究发现:在导叶开度9.8°下各工况的效率与扬程均有一定改善,额定工况点效率值提高0.56%,扬程提高2.10%;17.5°开度的额定工况点效率提高0.55%,扬程则提高0.018%,并且高效区得到一定拓宽;24.8°开度除小流量外其余工况点效率值与扬程均有明显提高;优化方案水泵水轮机泵工况内部流动特性得到改善.

In order to improve the pump-turbine′s pumping efficiency with long and short blade structure during operation,the partical swarm optimization(PSO) algorithm was improved by linearly changing inertial weights and changing learning factors,and then the improved PSO algorithm was used to optimize the pump turbine structure.A combination of test and numerical simulation was adopted,and the improved PSO algorithm of the learning factor was used to optimize the nine structural parameters to improve the pump turbine working condition efficiency and head. It is founded that under the guide vane opening of 9.8°,each working condition′s efficiency and head have been improved to a certain extent. The rated working point′s efficiency value is increased by0.56%,and the head is increased by 2.10%,with the rated working point efficiency of 17.5° increased by 0.55%,lifted by 0.018%.And the high-efficiency area is broadened to a certain extent. Except for small flow rate,the efficiency value and head of the other operating points of the 24.8° opening are significantly improved.