基于粒子群算法的多级低压涡轮一维气动优化设计方法

An Optimal Method of One-Dimensional Design for Multistage Low Pressure Turbines Based on Particle Swarm Optimization

关键词:  低维优化  多级低压涡轮  一维设计  子午流道  粒子群优化算法  改型设计  气动性能 

Key words:  Low dimensional optimization  Multistage low pressure turbines  One-dimensional design  Meridional channel  Particle swarm optimization  Improvement design  Aerodynamic performance

摘要:

为探究涡轮高效设计技术,从低维优化层面出发,提出了一种基于粒子群优化算法的多级低压涡轮一维设计和优化方法。该方法以涡轮效率为目标,通过建立涡轮子午流道形式以及气动特性等约束条件将涡轮一维设计转化成包含约束限制的极大值优化问题。在验证粒子群算法优化性能的基础上发展了多级低压涡轮一维气动优化设计程序,该程序通过优化地选取涡轮各级的多个设计变量,有效地生成满足多个约束条件的级最佳速度三角形以及最佳初步子午流道形式。利用该程序完成了原型低压涡轮的优化改型工作,三维数值模拟结果表明,优化改型方案在设计点和非设计点的气动性能均获得了不同程度的改善。 

Abstract:

In order to explore an efficient designing technique for turbines, a method of one-dimensional design for multistage low pressure turbines was presented, from the view of low dimensional optimization. Based on the particle swarm optimization(PSO) theory,the method transfers the one-dimensional design into constrained maximum optimization by taking the efficiency of the turbine as objective and a series of functions concerning the meridional channel and aerodynamic characteristics as constraints. The PSO was verified effectively based on which a program for optimized design of multistage low pressure turbines was developed. By selecting some design parameters of each turbine stage optimally, this program can generate an optimized velocity triangle of each stage and a preliminary meridional channel that satisfies the constraint conditions effectively. Furthermore, an improvement design of an original low pressure turbine was accomplished using the program. Based on the three dimensional simulation, this optimized scheme improves the performance to varying degrees in terms of design point and off-design point. 