阴极
电解
法律工程学
工程类
汽车工程
废物管理
电气工程
化学
电极
电解质
物理化学
作者
Francisco J. delaCalle,A. Blanco,Daniel F. García,Rubén Usamentiaga
标识
DOI:10.1109/tia.2024.3384142
摘要
The current steel industry needs new efficient monitoring systems that to aid human operators in their tasks. One of the most dangerous steps in the electrolysis process is the moment in which the cathodes and the anodes are lifted from the cells and moved through the facility. This paper proposes a new inspection system to detect whether the cathodes can be lifted by the crane. The proposed system uses a camera that takes images of the rack of cathodes before they are lifted. The images are processed to detect incorrectly positioned cathodes that could generate a problem when lifted. The computer vision system used to process the images was tested on more than 5500 cathodes with 90% effectiveness on anomaly detection and 0.8% of detection errors. The results were then validated in a second industrial facility on more than 6000 cathodes providing even better results, proving the effectiveness and reliability of the system. The system has been also tested in terms of response time proving it suitable for real-time monitoring in a real facility 1 .
科研通智能强力驱动
Strongly Powered by AbleSci AI