テクノロジーと最適化のグローバル ジャーナル

Prediction of weld area based on image recognition and machine learning

Abstract

Milorad Bojic

Modern aluminium alloy welding techniques like laser oscillation welding can successfully reduce weld porosity brought on by the physical and chemical characteristics of aluminium alloy. Since it has a significant impact on the mechanical qualities of welded connections, the weld area is frequently used as an evaluation index of geometric attributes to assess the welding quality. In this paper, a method for predicting the weld area for laser oscillation welding of 6061 aluminium alloy is proposed. The cross-sectional area of the weld is computed using image recognition technology from the metallographic micrographs of welding trials, and the inaccuracy of the recognised weld area is less than 8.8%. Additionally, alternative prediction models for the weld area are created by machine learning methods, such as linear regression, under varied process circumstances.