基于人工神经网络和遗传算法的动叶可调轴流风机后导叶数值优化  

Numerical optimization for outlet guide vane of an adjustable rotor blade axial fan based on artificial neural network and genetic algorithm

作　　者：陈得胜 白新奎 张振民 马翔 CHEN Desheng;BAI Xinkui;ZHANG Zhenmin;MA Xiang(Xi’an Thermal Power Research Institute Co.,Ltd.,Xi’an 710054,China;Huaneng Gansu Energy Development Co.,Ltd.,Lanzhou 730070,China;School of Energy and Power Engineering,Xi’an Jiaotong University,Xi’an 710049,China)

机构地区：[1]西安热工研究院有限公司,陕西西安710054 [2]华能甘肃能源开发有限公司,甘肃兰州730070 [3]西安交通大学能源与动力工程学院,陕西西安710049

出　　处：《热力发电》2021年第10期142-149,共8页Thermal Power Generation

基　　金：西安热工研究院有限公司研发基金项目(GA-18-TZK010)。

摘　　要：以自主设计的单级动叶可调轴流风机模型为研究对象,建立了基于人工神经网络和遗传算法的三维气动优化设计模型,在设计点对后导叶进行三维气动优化生成优化叶型,并运用全通道三维数值计算方法对后导叶优化前后的风机整体性能进行数值模拟。模拟结果表明:与原始叶型相比,设计点优化叶型风机效率提高了1.1%,风机全压提高了2.1%;非设计点风机效率提高了8.6%,风机全压提高了20.3%。通过进一步对叶轮级内部流场分析可知,无论是在设计点还是在非设计点,优化叶型较原始叶型不仅表现了更强的抗气流分离能力,而且能够有效抑制叶栅通道内低能流体团的径向迁移,因而优化叶型显著减少了叶栅通道内部的流动损失。Taking a self-designed single-stage adjustable rotor blade axial flow fan model as the research object,a three-dimensional aerodynamic optimization design model based on artificial neural network(ANN)and genetic algorithm(GA)is established.The 3D aerodynamic optimization for outlet guide vane is carried out to generate optimized blade profile at the design point,and the overall performances of the fan before and after the optimization of the outlet guide vane are numerically verified using the full-channel 3D numerical calculation method.The simulation results show that,compared with the original blade profile,for the optimized blade profile at design point,the efficiency value of the fan increases by 1.1%,and the total pressure value of the fan increases by 2.1%.For the optimized blade profile at off-design point,the efficiency value of the fan increases by 8.6%,and the total pressure value of the fan increases by 20.3%.Through further analysis on internal flow field of the impeller stage,conclusions can be made that whether it is at the design point or at the off-design point,the optimized blade profile not only can exhibit stronger resistance to airflow separations than the original blade profile,but also can effectively suppress the radial migration of low-energy fluid in the cascade channel,and thus the optimized blade profile significantly reduces the flow losses inside the cascade channel.

关 键 词：动叶可调轴流风机 人工神经网络 遗传算法 后导叶 叶型优化 风机效率 

分 类 号：TK223.26[动力工程及工程热物理—动力机械及工程]