基于BP神经网络和NSGA-Ⅱ的离心压气机机匣处理槽参数优化  

Parameters Optimization of Centrifugal Compressor Casing Slot Based on BP Neural Network and NSGA-Ⅱ

作　　者：陈金萍 叶顺宏 李颂 CHEN Jinping;YE Shunhong;LI Song(School of Applied Technology,Dalian Ocean University,Dalian 116300,China;School of Mechanical Engineering,Liaoning Technical University,Fuxin 123000,China;Datong North Tianli Supercharging Technology Co.,Ltd.,Datong 037036,China)

机构地区：[1]大连海洋大学应用技术学院,辽宁大连116300 [2]辽宁工程技术大学机械工程学院,辽宁阜新123000 [3]大同北方天力增压技术有限公司,山西大同037036

出　　处：《车用发动机》2023年第1期62-68,共7页Vehicle Engine

基　　金：辽宁省自然科学基金项目(2019-MS-160)。

摘　　要：为获取离心压气机机匣处理槽的最优结构参数,针对某型离心压气机机匣处理的槽参数展开了优化工作,通过已有模拟数据建立BP神经网络预测模型,利用遗传算法(NSGA-Ⅱ)对槽的结构参数进行了多目标寻优工作。结果表明:优化后的结构参数为开槽宽度4 mm,槽中心位置为靠近叶轮前缘距离导风轮中段1.5 mm处。经过模拟分析,优化值对应的槽处理结构位置相对靠近叶轮前缘,使低速区域相对前移,改善了主通道内的流动情况,喘振边界在高转速下明显向小流量偏移,进一步拓宽了压气机稳定工作范围。In order to obtain the optimal structural parameters of a centrifugal compressor casing treatment slot,the optimization of centrifugal compressor casing slot parameter was carried out.A BP neural network prediction model was established based on the existing simulation data,and a multi-objective optimization of the slot structural parameters was carried out by using the genetic algorithm(NSGA-Ⅱ).The results show that the optimized structural parameters are the slot width of 4 mm and the slot center position of 1.5 mm from the middle section of air guide wheel near the leading edge of impeller.According to the simulation analysis,the slot treatment structure corresponding to the optimized value is relatively close to the leading edge of impeller,which makes the low-speed region relatively forward and improves the flow situation of main channel so as to shift the surge boundary obviously to the small flow rate at high speed,the stable working range of compressor widens.

关 键 词：参数优化 离心压气机 BP神经网络 遗传算法 数值模拟 

分 类 号：TH452[机械工程—机械制造及自动化]