摘要
针对薄壁齿圈的装夹变形问题,将Abaqus有限元仿真与BP神经网络技术应用到了齿圈装夹变形预测中。根据齿圈实际加工装夹情况,应用Abaqus有限元分析软件,建立了齿圈装夹变形的仿真模型,开展了齿圈装夹变形的有限元分析研究,建立了齿圈装夹力及其径向最大装夹变形之间的关系;以Abaqus有限元仿真数据作为训练样本和检验样本,借助BP神经网络良好的预测精度和非线性泛化能力,通过MATLAB神经网络工具箱,建立了基于BP神经网络的齿圈装夹变形预测数字化模型;并根据检验样本对模型进行了检验,预测值与仿真值之间的相对误差在0.05%之内。研究结果表明:建立的基于BP神经网络的齿圈装夹变形预测数字化模型是准确有效的,可以为智能化大数据加工制造环境下的齿圈装夹参数优化提供准确有效的数据。
Aiming at the problem of clamping deformation of thin-walled ring gear,Abaqus finite element simulation and BP neural network technology were applied to the prediction of gear ring deformation.According to the actual machining and clamping of the gear ring,the Abaqus finite element analysis software was used to establish the relathonsilp between the simulation model of the gear ring clamping deformation,and the finite element analysis of the ring gear clamping deformation was carried out to establish the relationship between the ring gear clamping force and its radial maximum clamping deformations.Based on Abaqus finite element simulation data as training samples and test samples,with the good prediction accuracy and nonlinear generalization ability of BP neural network,BP neural network based digital model of gear ring deformation prediction was established by MATLAB neural network toolbox,and the model was tested according to the test sample,and the relative error between the predicted value and the simulated value was within 0.05%.The results indicate that the established BP neural network based digital model of gear ring deformation prediction is accurate and effective,and can provide accurate and effective data for the optimization of gear ring clamping parameters in intelligent big data processing and manufacturing environment.
作者
韩军
张磊
段荣鑫
王静
HAN Jun;ZHANG Lei;DUAN Rong-xing;WANG Jing(School of Mechanical Engineering,Inner Mongolia University of Science and Technology,Baotou 014010,China)
出处
《机电工程》
CAS
北大核心
2020年第6期641-646,共6页
Journal of Mechanical & Electrical Engineering
基金
内蒙古自治区高等学校科学研究项目(NJZZ19125)。
关键词
齿圈
有限元仿真
BP神经网络
装夹变形预测
gear ring
finite element simulation
BP neural network
clamping deformation prediction
