Use of the Hough transformation to detect lines and curves in pictures
Communications of the ACM
Uncertainty modelling using Dempster-Shafer theory for improving detection of weld defects
Pattern Recognition Letters
Software Architecture in Practice
Software Architecture in Practice
IEEE Transactions on Software Engineering
IJCAI'83 Proceedings of the Eighth international joint conference on Artificial intelligence - Volume 2
| Hi-index | 0.00 |
Assessing the quality of a weld in the steelmaking industry is a complex task. The level of complexity increases when the assessment is based on non-destructive tests. Skilled technicians are often required to make a decision based on automatic assessments of welds. Technicians consider the results of the automatic assessments and use their expert knowledge in order to make a final decision about the quality of the weld. In this paper we propose a decision support system to assess the quality of resistance seam welds of steel strips based on statistical analysis of both the mechanical and electrical variables involved in the welding process to be assessed as well as previously recorded historical data of similar welds. The proposed system is designed following component model based software architecture. The system consists of a set of orthogonal modules: welding variable measurement, welding variable processing and welding quality assessment, communicated by means of dedicated interfaces. The proposed system has been installed in three steel manufacturing lines. With the reduction in the time spent by technicians to make a decision about each weld, the productivity of the manufacturing line has greatly improved. Furthermore, production costs have been reduced since the number of defective welds assessed as non-defective was reduced, and thus the failures in the manufacturing lines due to weld breakages. The experimental results after two years of use in a steel strip galvanizing line are shown.