基于主轴动态特性的GA-BP网络预测FSW接头性能

张天鹏; 赵亚东

doi:10.15980/j.tzzz.2022.07.005

研究·计算机应用 | 浏览量 : 0 下载量: 59 CSCD: 0

PDF
导出
分享
收藏
专辑

基于主轴动态特性的GA-BP网络预测FSW接头性能
GA-BP Neural Network Prediction of FSW Joint Properties Based on Dynamic Characteristics of Spindle
2022年42卷第7期页码：814-818
纸质出版日期： 2022 ，
DOI： 10.15980/j.tzzz.2022.07.005

扫描看全文

张天鹏, 赵亚东. 基于主轴动态特性的GA-BP网络预测FSW接头性能[J]. 特种铸造及有色合金, 2022,42(7):814-818.

Zhang Tianpeng, Zhao Yadong. GA-BP Neural Network Prediction of FSW Joint Properties Based on Dynamic Characteristics of Spindle[J]. Special Casting & Nonferrous Alloys, 2022,42(7):814-818.
张天鹏, 赵亚东. 基于主轴动态特性的GA-BP网络预测FSW接头性能[J]. 特种铸造及有色合金, 2022,42(7):814-818. DOI： 10.15980/j.tzzz.2022.07.005.

Zhang Tianpeng, Zhao Yadong. GA-BP Neural Network Prediction of FSW Joint Properties Based on Dynamic Characteristics of Spindle[J]. Special Casting & Nonferrous Alloys, 2022,42(7):814-818. DOI： 10.15980/j.tzzz.2022.07.005.

摘要

以搅拌摩擦焊过程中主轴的动态特性(主轴力和振动加速度)为输入参量，以接头的抗拉强度为输出参量，构建了4×8×13×1四层拓扑结构的遗传算法-BP神经网络模型，对1060铝/AZ31B镁异种材料搅拌摩擦的对接焊接头的力学性能进行预测。分别选取37组和8组样本数据对模型进行训练和测试。结果表明，训练样本和测试样本的预测值与实测值的平均相对误差分别为4.1%和5.4%

均在可接受范围内，表明该模型具有较高的预测精度。

Abstract

Taking the dynamic characteristic（force and vibration acceleration） of the spindle during the friction stir welding process as input parameters

and tensile strength of the joint as output parameters

a genetic algorithm BP neural network model with 4×8×13×1 four-layer topology was established to predict the mechanical properties of 1060 Al/AZ31 B Mg dissimilar metals friction stir butt welding joints.37 and 8 groups of sample data were selected respectively to train and test the model. The results reveal that the average relative errors between the predicted values and the measured values of the training samples and the test samples are within the acceptable range

which are 4.1% and 5.4%

respectively

indicating the high prediction accuracy of the model.