Data-driven feature selection for multisensory quality monitoring in ARC welding

Abstract

Feature selection is the key issue for multisensory data fusion-based online welding quality monitoring in the area of intelligent welding process. This paper mainly focus on the automatic detection of typical welding defect for Al alloy in gas tungsten arc welding (GTAW) by means of a series of analysis of synchronous online arc spectrum, arc sound pressure and arc voltage signal. Based on the developed feature selection algorithms, hybrid fisher-based filter and wrapper was successfully utilized to evaluate the sensitivity of each feature and reduce the feature dimensions. Finally, the optimal feature subset with 19 features was selected to obtain the highest accuracy, i.e., 94.72 % of the established classification model support vector machine-cross validation (SVM-CV).