基于敏感性分析降维的多翼离心风机叶轮优化设计  

Optimal Design of Squirrel-cage Fan Impeller Based on Sensitivity Analysis and Dimension Reduction

作　　者：王镇 李震[1] 吴兰勇 解博超 石清鑫[3] 张楚华[1] WANG Zhen;LI Zhen;WU Lanyong;XIE Bochao;SHI Qingxin;ZHANG Chuhua(School of energy and power engineering,Xi'an Jiaotong University,Xi'an 710049;Guangdong Green Island Air System Co.,Ltd,Guangzhou 1529200;Xian Thermal Power Research Institute Co.,Ltd,Xi'an 710049)

机构地区：[1]西安交通大学能源与动力工程学院,西安710049 [2]广东绿岛风空气系统股份有限公司,广州529200 [3]西安热工研究院有限公司,西安710049

出　　处：《工程热物理学报》2023年第2期397-405,共9页Journal of Engineering Thermophysics

基　　金：国家科技重大专项(No.2017-Ⅱ-0006-0020);机械结构强度与振动国家重点实验室项目(No.SV2021-ZZ-23);国家自然科学基金项目(No.51776154)。

摘　　要：针对多翼离心风机内部流动分离严重、流动损失大及叶轮几何参数多等特点,本文以风机叶轮为对象,开展了几何参数化建模及基于敏感性分析降维,并在此基础上,通过建立包含拉丁超立方抽样法、支持向量回归(SVR)代理模型、混合自适应采样方法和遗传算法的高维高效优化方法,分别开展了效率单目标优化与效率–全压多目标优化设计。研究结果表明:对于风机效率,几何参数影响程度由大到小依次为叶轮内径、叶片出口角、叶轮外径、中盘位置、叶片进口角、叶片数、叶轮宽度、叶片厚度。单目标优化后风机最高效率点效率提高2.29%;多目标优化后效率提高1.91%,全压提升30.95 Pa。研究工作对于认识多翼离心风机内流机理、发展先进优化设计方法具有较重要的学术意义及工程应用价值。In view of the characteristics of severe internal flow separation, large flow loss and multiple impeller geometric parameters of squirrel-cage fan, this paper takes the fan impeller as the object to carry out the geometric parametric modeling and dimension reduction based on sensitivity analysis. On this basis, the high-dimensional and efficient optimization methods including Latin Hypercube sampling method, support vector regression(SVR) surrogate model, mixed adaptive sampling algorithm and genetic algorithm are established to carry out the total efficiency single-objective optimization and total efficiency-total pressure rise multi-objective optimization design,respectively. The results show that for total efficiency, the influence degree of geometric parameters from large to small is the impeller inlet diameter, the blade outlet angle, the impeller outlet diameter,the annular plate position, the blade inlet angle, the blade number, the impeller width and the blade thickness. The total efficiency of best efficiency point is increased by 2.29% after single-objective optimization. After multi-objective optimization, the total efficiency and total pressure rise are increased by 1.91% and 30.95 Pa respectively. The research has important academic significance and engineering application value for understanding the internal flow mechanism of squirrel-cage fan and developing advanced optimization design method.

关 键 词：多翼离心风机 优化设计 遗传算法 代理模型 敏感性分析 

分 类 号：TH43[机械工程—机械制造及自动化]