The management of industrial arc welding by neural networks

New methods of monitoring industrial process variables are constantly being sought with the aim to improve control efficiency. It has been observed that skilled welders subconsciously adapt their manual arc welding technique in response to a variation in the sound produced from the process. This paper proposes an approach to the control of an automated submerged arc welding process using:- 1. Real time monitoring of acoustic emissions 2. The application of neural networks to predict the point of instability of the process variables.