Use of Machine Learning Algorithms for Weld Quality Monitoring using Acoustic Signature

摘要: Abstract Welding is one of the major joining processes employed in fabrication industry, especially that manufactures boiler, pressure vessels, marine structure etc. Control weld quality very important for such industries. In this work an attempt made to correlate arc sound with quality. The welding done various combinations current, voltage, and travel speed produce good welds as well defects. defects considered study are lack fusion burn through. Raw data points captured from were converted into amplitude signals. welded specimens inspected classified 3 classes Statistical features raw extracted using mining software. Using classification algorithms classified. Two namely, J48 random forest used efficiencies reported.