舰用钢板的大功率激光焊接参数优化研究

发布时间：2018-01-22 01:44

本文关键词： 激光焊接双权值神经网络声信号特征提取　出处：《电子技术应用》2017年02期 　论文类型：期刊论文

【摘要】：针对高强低碳合金厚钢板大功率激光焊接过程的非线性、多变量耦合性、不确定性等特点,通过构建基于声信号采集的监测系统,实现舰用钢板焊接过程声音的实时采集,并提出特征向量,构建双权值神经网络(DWNN)模型,充分利用DWNN优秀的非线性拟合能力,实现大功率激光焊接多参数与声信号多特征之间非线性映射的神经网络建模。在拟合精度和迭代次数上,DWNN比径向基函数网络等传统网络更优,为高强度低合金厚钢板的大功率焊接参数的优化和控制提供了良好的基础。
[Abstract]:Aiming at the characteristics of nonlinear, multi-variable coupling and uncertainty in the process of high-strength and low-carbon alloy thick steel plate laser welding, a monitoring system based on acoustic signal acquisition is constructed. The real-time acquisition of the sound in the welding process of ship steel plate is realized, and the eigenvector is put forward, and the dual weight neural network (DWNN) model is constructed to make full use of the excellent nonlinear fitting ability of DWNN. The neural network modeling of nonlinear mapping between multi-parameter and acoustic signal in high-power laser welding is established. DWNN is better than traditional network such as radial basis function network in fitting accuracy and iteration times. It provides a good basis for the optimization and control of high power welding parameters of high strength and low alloy thick steel plate.
【作者单位】：南通大学电气工程学院;
【基金】：国家自然科学基金(61273024,61305031) 江苏省自然科学基金(BY2016053-11) 江苏省“333”高层次人才培养工程(BRA2015366)
【分类号】：TG456.7
【正文快照】： 0引言国内外对舰用高强低合金钢的主要焊接方式中间工序繁杂,能量消耗巨大,机器人难以胜任[1]。激光电弧复合焊接技术可以进一步地提升焊接速度,增加焊接材料的厚度,加强间隙桥接能力,极大地提高效率[2]。焊接过程由于焊条裂化和熔融金属振动而伴随着声音,基于麦克风的焊接

【相似文献】