基于BP人工神经网络模型的激光焊缝宽度预测方法
Laser Weld Widths Prediction Based on BP Artificial Neural Network Model
王 琦1, 冯晓天2
点击:3804次 下载:0次
作者单位:(1. 镇江高等专科学校,江苏 镇江 212003; 2. 江苏大学 机械工程学院,江苏 镇江 212013)
中文关键字:焊缝宽度; 激光拼焊板; 工艺参数; BP人工神经网络
英文关键字: weld width; LWTB; process parameters; BP artificial neural network
中文摘要:激光焊缝宽度是考核激光拼焊板质量的重要指标之一,直接影响到拼焊板的成形性能。因此,通过对激光焊缝宽度进行预测可以达到焊接工艺参数优化的目的,以提高拼焊板的焊接质量与成形性能。本文利用BP人工神经网络技术建立了焊缝宽度预测模型,该模型可以实现对焊缝宽度的有效预测,预测精度达到96%以上,具有较好的工业实用价值。
英文摘要:Laser weld widths is one of the important quality assessment indicators about laser tailor-welded blank(LWTB), which directly affects LWTB forming properties. In order to improve the welding quality and forming properties, process parameters were optimized by laser weld widths prediction. Weld widths prediction model with good prediction effect was put forward by means of BP artificial neural network technology. The forecast accuracy of the weld widths were more than 96%. Therefore, the proposed model has good practical value in industrial.