中文摘要：

以某双排对转压气机为研究对象,采用三维Navier-Stokes方程计算、叶片参数化造型、神经网络构建近似函数、遗传算法寻优相结合的优化方法,对转子2叶片进行级环境下的全三维叶型优化方案探索.优化目标是在控制流量和压比的情况下,最大化等熵效率.优化结果显示,设计点压气机效率提高1.5%左右.通过流场细微结构分析表明新叶型的气动布局得到合理改善,转子2中,尖部流场低速区域有所减少,分离现象有所减弱.

英文摘要：

With the optimization platform of three-dimensional（3-D） design,the optimization design was carried out for rotor 2 of a two stage counter-rotating compressor in multistage environment based on the methods such as： three-dimensional Navier-Stokes（N-S） flow computation,blade section parameterization,artificial neural networks and genetic algorithm.The optimization objective is to maximize the isentropic efficiency while the mass flow and total pressure ratio are controlled.Numerical results show that efficiency is increased by 1.5% at design point after optimization of rotor 2.The gain in aerodynamic performance is largely attributable to the reduction of losses due to flow separation in suction side,as observed through flow field analysis.