| 2002 №01 (05) | 2002 №01 (07) |
The Paton Welding Journal, 2002, #1, 44-46 pages
Adaptive algorithm of quality control of resistance spot welding using neural networks
N.V. Podola, P.M. Rudenko, V.S. Gavrish
E.O. Paton Electric Welding Institute of the NASU 11 Kazymyr Malevych Str., 03150, Kyiv, Ukraine.Abstract
Adaptive algorithm is suggested for quality control of the resistance spot welding of low-carbon galvanized steel taking into account the change in thickness of parts being welded. Algorithm was verified experimentally for (0.8 + 0.8), (1.2 + 1.2) and (2.0 + 2.0) mm packs. The adaptation of control algorithm from the results of spot welding of a limited number of samples provides an error in prediction of nugget diameter using a neural network of not more than 10 %.
Keywords: resistance spot welding, diameter of spot nugget, quality control, neural networks, adaptation
References
1. Paton, B.E., Podola, N.V., Gavrish, V.S. et al. (1998) Automatic evaluation of quality of resistance spot welding using neuron networks. Avtomatich. Svarka, 12, 3-10.2. Pustyinik, E.I. (1986) Statistic methods of analysis and processing of observations. Moscow: Nauka.