447 音声による NC 工作機械の異常検出法

概要

論文の詳細を見る
This paper describes a monitoring system for NC machine tool at remote place through computer network. Most conventional methods use parameters like spindle speed, feed speed, etc. and the operator needs to watch them continuously. In this paper, we propose a semi-automatic trouble detection method without using the above-mentioned parameters. Sound and video of the operation are continuously collected and only the video is sent to the operator. The system detects the trouble automatically by performing frequency analysis of the sound using Fast Fourier Transform (FFT). The essence of semi-automatic trouble detection system here is that even if the system detects any trouble in the NC machine, it does not stop the operation, but sends an emergency signal to the operators. It is the operator who decides on stopping or continuing the operation by checking the video signal.
一般社団法人日本機械学会の論文
2000-09-01