Welding and weaving speed estimation using acceleration sensor based on SVM

Abstract

The amount of heat applied to the welding part by the movement of the welding torch has a predetermined standard per unit area, according to the base material and the welding method. When the heat input is large, the cooling rate is slowed and the ductility is increased; however, as the hardness is decreased, the welding part is easily deformed by the residual stress. Further, if the amount of heat input is small, the weld zone tends to separate due to the lack of welding. Therefore, for a constant amount of heat to be applied to the weld, it is essential to measure the real-time moving speed, direction, and angle of the torch during welding. In this paper, we propose welding and weaving speed estimation system using an acceleration sensor based on the support vector machine (SVM). To measure the welding and weaving speeds, we analyze the obtained signal from the acceleration sensor in the frequency domain, based on the fast Fourier transform (FFT), and derive the feature vector. Based on this, we apply the SVM to improve the accuracy of low speed measurement by classifying the characteristics of the measurement signal according to the speed change. To verify the validity, we analyze the accuracy and error of the proposed algorithm according to the change in welding and weaving speed. The simulation result shows that the proposed method achieves an average positioning accuracy of 93.36% and a maximum positioning error of 0.3834 m.