摘要
为提高离心压缩机气动性能,考虑二元叶轮优化多以构型参数为变量,不适用于变量间存在交互作用的三维叶片。以某三元叶轮为优化对象,考虑各变量间的交互作用,采用Bézier曲线加载叶片角,直接选取三维叶片型线为优化变量进行气动优化。结合CFD数值仿真、非线性映射能力较强的BP神经网络与全局寻优能力较强遗传算法,以提升等熵效率和压比为目标进行优化。结果表明:设计工况下,优化后的叶轮等熵效率同比提升1.9%,压比同比提升0.3。三维叶片中部向吸力面偏移,叶片尾缘向压力面偏转可同时提升叶轮的效率和压比。
In order to improve the aerodynamic performance of centrifugal compressor,the optimization of binary impeller is mainly based on configuration parameters,which is not suitable for three-dimensional blades with interaction between variables.Taking a three-way impeller as the optimization object,considering the interaction between variables,Bézier curve was used to load the blade Angle,and three-dimensional blade profile was directly selected as the optimization variable for aerodynamic optimization.Combined with CFD numerical simulation,BP neural network with strong nonlinear mapping ability and genetic algorithm with strong global optimization ability,the optimization is carried out to improve the isentropic efficiency and pressure ratio.The results show that under design conditions,the isentropic efficiency of the optimized impeller increases by 1.9%and the pressure ratio increases by 0.3 compared with the preceding impeller.The efficiency and pressure ratio of the impeller can be improved simultaneously when the middle of the blade is shifted to the suction surface and the trailing edge of the blade is deflected to the pressure surface.
作者
张春峰
赵海峰
张铁柱
ZHANG Chun-feng;ZHAO Hai-feng;ZHANG Tie-zhu(School of Chemical Equipment,Shenyang University of Technology,Liaoyang 111000,China)
出处
《汽轮机技术》
北大核心
2023年第1期20-22,76,共4页
Turbine Technology
关键词
离心压缩机
叶片型线
叶片参数化
BP神经网络
遗传算法
centrifugal compressor
leaf blade type line
blade parameterization
BP neural network
genetic algorithm