摘要
为了预测变速箱总成密封质量,提高一次装配合格率,提出一种灰关联分析和遗传算法-BP神经网络的预测模型。应用灰关联分析方法,计算各装配工艺参数的灰关联度,确定影响变速箱总成密封质量的关键装配工艺参数。建立基于遗传算法GA的BP神经网络预测模型,对M型变速箱装配总成泄漏值进行预测。几种预测方法结果表明,本文提出的预测方法获得的预测值平均相对误差最小,约为5.67%,验证了该预测方法的有效性。
In order to predict the seal quality of the gearbox assembly and improve the first-time assembly pass rate,a grey relational analysis and genetic algorithm-BP neural network prediction model is proposed.Applying the grey relational analysis method,the grey relational degree of the assembly process parameters is calculated,and the key assembly process parameters that affect the sealing quality of the gearbox assembly are determined.A BP neural network prediction model based on GA is established to predict the leakage value of M-type gearbox assembly.The results of several prediction methods show that the average relative error of the prediction value obtained by the proposed prediction method is the smallest,about 5.67,which verifies the effectiveness of the prediction method.
作者
周康渠
张朝武
屈清
唐蔗湛
ZHOU Kangqu;ZHANG Chaowu;QU Qing;TANG Zhezhan(College of Mechanical Engineering,Chongqing University of Technology,Chongqing 400054,China;Qingling Motors Co.Ltd,Chongqing 400052,China)
出处
《工业工程》
北大核心
2022年第2期22-27,41,共7页
Industrial Engineering Journal
基金
重庆市技术创新与应用示范重点示范项目(cstc2018jszx-cycdX0169)。
关键词
变速箱
密封质量
预测
灰关联分析
遗传算法
BP神经网络
gearbox
seal quality
prediction
grey correlation analysis
genetic algorithm
BP neural network
