A Modified Blind Signal Extraction Approach for the Spot Welding Quality Monitor System

摘要

In this paper, a Spot Welding Quality Monitor System based on sensors array technique is applied in the course of the quality detecting. In order to use the monitor system more efficiently, the experiment in this study is designed into two parts. One part is the scheme selecting of the linear magnetic sensors array. The other is choosing a perfectible case to detect the process of the aluminum alloy spot welding and then using the modified FastICA algorithm to extract the objective signals. The results showed that, under the selected parameters, the Spot Welding Quality Monitor System works well during the welding process, and the magnetic signal source can be extracted validly with the modified algorithm during the welding process.