基于均匀设计的粒子群优化在减压化设计中的应用

采用粒子群优化算法,对涡轮泵气封减压试验系统中的关键部件减压器进行结构优化设计.在优化过程之前采用两次均匀设计的方法缩小设计变量的取值范围,以提高计算效率.通过调节气封系统中的主要部件气体减压器的结构参数,使得减压器的出口压强稳定,满足工程需求.优化结果显示,通过两次均匀设计后,设计变量的取值范围得到有效压缩,减少了优化过程的计算量.与初始设计相比,减压器的出口压强振幅明显减弱,达到了工程设计的要求.

Particle swarm optimization of pressure reducing regulator based on uniform design

The pressure reducing regulator which was the key point of turbo-pump gas-sealed decompressing experiment system was optimized by PSO(particle swarm optimization).Uniform design was used twice before the optimization for the sake of reducing computing quantity.The structure parameters of pressure reducing regulator were optimized to obtain steady outlet pressure meeting engineering requirement.The result indicates that the range of variables was compressed availably through twice of the uniform design.Compared with initial point,the swing of outlet pressure decreases obviously according with engineering design.