摘要
三维气动优化设计方法是提高叶轮机械性能的有效手段.应用三维优化方法,对一涡轮增压器的径流式涡轮叶片进行了叶型优化设计.优化设计使用神经网络与遗传算法相结合的优化算法;变化控制涡轮端壁型线、积叠规律、中弧线和厚度的自由参数得到不同的叶轮几何,这些几何通过气动计算得到的性能作为神经网络优化的输入参数,以总-静效率为优化目标进行优化.优化后的涡轮总静效率比原涡轮提高了1.5%.这些工作为径流涡轮叶轮的多学科优化设计提供良好的基础.
The 3D aerodynamic optimization design method is an effective way to improve performance of turbomachinery. Optimization to a turbocharger radial turbine impeller base on this method is carried out. The optimization is performed by using artificial neural network （ANN） and genetic algorithm. Different blade geometric parameters are obtained by controlling such free variables as end-walls, stacking laws, camber curves and thickness. Performance from aerodynamic simulation with the blade parameters are used as input of ANN, and a total-to-static efficiency is set as the target in the optimization. The performance of optimized blade is increased 1.5% comparing to the original impeller. The study provides a useful design reference for muhidisciplinary optimization of radial turbine.
出处
《车辆与动力技术》
2009年第4期1-5,共5页
Vehicle & Power Technology
关键词
径流涡轮
气动计算
优化设计
radial turbine
aerodynamic simulation
optimization design
