Conveyor Belt Vibration Analysis Using High-Speed Video Images

Abstract

This study proposes a novel video-based conveyor monitoring method that can estimate distances between belt-supporting rollers by analyzing belt vibration responses in high-frame-rate (HFR) images. Focusing on the fact that the vibration components of a belt include frequencies inversely proportional to the distance between its ends, roller-to-roller distances that correspond to the state of contact between a belt and rollers are estimated by HFR-video-based analysis that can detect vibration peak frequencies at multiple belt positions without directly observing them captured in images. Its effectiveness was verified by conducting a wide-area monitoring test of a material handing conveyor in ironworks for roller-to-roller distance estimation with vibration frequency analysis at multiple points on its several-tens-of-meters belt in 9344×600 images captured at 300 fps.