摘要
提出了一种基于DeepLab-EMCAD的全位置焊接熔池识别与偏差测量方法。采用轻量级MobileNetV3网络作为模型编码器主干网络,并优化空洞空间金字塔池化模块,以减少模型参数量、提高分割效率。在解码器中集成EMCAD多注意力机制,提高熔池的分割精度。基于熔池分割结果,提出一种偏差计算方法来定量描述偏差。实验结果显示,相较于基线模型,所提出的模型在熔池分割中的平均交并比、平均像素精确度分别提高了5.72%和5.5%,推理时间缩短了29.69 ms,参数量减少了4.854×10^(7)个;相比于经典分割网络,所提出的模型在熔池边缘处理上效果更优,焊接偏差检测误差控制在0.1 mm之内。
A method for full-position welding pool identification and deviation measurement was proposed based on DeepLab-EMCAD.A lightweight MobileNetV3 network was adopted as the backbone of the model encoder,and the atrous spatial pyramid pooling(ASPP)module was optimized to reduce the model parameters and improve the segmentation efficiency.The EMCAD multi-attention mechanism was integrated into the decoder to enhance the segmentation accuracy of the welding pools.A deviation calculation method was proposed to quantitatively describe the deviation based on the segmentation results of the welding pools.Experimental results show that compared with the baseline model,the proposed model improves the average intersection over union and average pixel accuracy in welding pool segmentations by 5.72%and 5.5%respectively,and the inference time is reduced by 29.69 ms,with the number of parameters decreasing by 4.854×10^(7).Compared with classic segmentation networks,the proposed model has the best performance in handling the edges of the welding pools.The deviation detection errors are controlled within 0.1 mm.
作者
熊烨成
刘海生
王中任
史铁林
夏海
杨洪波
XIONG Yecheng;LIU Haisheng;WANG Zhongren;SHI Tielin;XIA Hai;YANG Hongbo(School of Mechanical Engineering,Hubei University of Arts and Science,Xiangyang,Hubei,441053;Xiangyang Key Laboratory of Intelligent Manufacturing and Machine Vision,Xiangyang,Hubei,441053;School of Mechanical Science and Engineering,Huazhong University of Science and Technology,Wuhan,430074)
出处
《中国机械工程》
北大核心
2025年第12期2993-3001,共9页
China Mechanical Engineering
基金
湖北省自然科学基金(2022CFD080)。
关键词
管道全位置焊接
高效多尺度卷积注意力解码器(EMCAD)
熔池
语义分割
偏差测定
pipe full position welding
efficient multi-scale convolutional attention decode(r EMCAD)
welding pool
semantic segmentation
deviation measurement
