Embedded Continuous Casting Slag Detection System Based on Artificial Neural Network
Aiming at the problems of molten steel continuous casting slag detection system (SDS), such as high cost,short working life etc, a kind of embedded SDS realization method based on artificial neural network (ANN) technology was put forward. Technical process detail of continuous casting was investigated and analyzed, vibration shock signal of molten steel was looked upon the main detecting signal. According to geometric similarity principle, embedded water model experiment platform was established combined embedded system technology. Using self-designed real-time data collection and analysis system to collect vibration signal and extract characteristic data, preprocessed data was trained and recognized by ANN, and molten steel status was identified to realize automatic control for continuous casting production according to corresponding control strategy. Industrial experiments results proved that this method requires low cost and little re-building for current devices, can realize pre-warning for slag carry-over, and its slag detection ratio can reach more 99%.
continuous casting slag detection vibration measurement embedded ANN
LI Pei-yu LI Ju-zhong CHEN Wen-cheng
Mechanical Engineering Department, Zhejiang University, Hangzhou 310027, Zhejiang, China Wuhan Iron and Steel Corporation, Wuhan 430080, Hubei, China Feng Hsin Iron & Steel Corporation, Taichung, Taiwan, China
国际会议
2009年薄板坯连铸连轧国际研讨会(TSCR 2009)(2009 International Symposium on Thin Slab Casting and Rolling)
南京
英文
541-550
2009-05-13（万方平台首次上网日期，不代表论文的发表时间）