基于并行多目标遗传算法大涵道分开式排气系统气动优化设计

通过引入快速非支配排序算法、拥挤距离以及拥挤距离比较算子等对基本遗传算法进行改进,并结合massage passing interface(MPI)并行编程技术,发展了主从式并行多目标遗传算法(PMGA).将PMGA与排气系统型面参数化设计方法、Navier-Stokes方程求解器相结合建立了分开式排气系统气动优化设计平台.应用该平台对某型分开式排气系统进行了多目标优化设计,得到了一组在三个目标上都优于初始设计的Pareto最优设计.将典型的Pareto最优设计和初始设计进行分析、比较,证明了该气动优化设计平台的高效性和可靠性.

Aerodynamic optimization design of high bypass ratio separate-flow exhaust system based on parallel multi-objective genetic algorithm

Simple genetic algorithm(GA) was improved with fast non-dominated sort approach,crowded distance estimation and crowded comparison operator.A new algorithm called parallel multi-objective genetic algorithm(PMGA) was developed using the master-slave parallel programming model with the support of massage passing interface (MPI).To establish the aerodynamic optimization design platform for high bypass ratio separate-flow exhaust system,PMGA was combined with profile parameterization of exhaust system and Navier-Stokes solver.With this platform,aerodynamic optimization design of the high bypass ratio separate-flow exhaust system was performed,and a set of Pareto-optimal solutions which was better than the initial design in three objectives was obtained.Detailed comparison and analysis of the Pareto-optimal designs and initial design confirmed the high efficiency and validity of the parallel multi-objective aerodynamic optimization design platform.