Steel water continuous casting slag detection system based on wavelet

被引：8

作者：

Tan, Dapeng ^{[1
]}

Li, Peiyu ^{[1
]}

机构：

[1] College of Mechanical and Energy Engineering, Zhejiang University

来源：

Jixie Gongcheng Xuebao/Chinese Journal of Mechanical Engineering | 2007年 / 43卷 / 02期

关键词：

Continuous casting; Feature extraction; Slag detection; Wavelet; Wavelet package;

D O I：

10.3901/JME.2007.02.141

中图分类号：

学科分类号：

摘要：

Based on lots of investigation and analysis for internal and overseas slag detection system (SDS) technology, a kind of SDS realization method using wavelet analysis is brought forward. According to the characteristic of design and realization for SDS, experiments platform for vibration measurement is established. After studying casting process from ladle to pouring box and protection casting supporting mechanical structure, vibration dynamics model of supporting structure is founded, also variation of steel water impact force and influencing factors of inspecting spot vibration are discussed. Primary analysis is done for steel water vibration signal by the way of basal wavelet processing method, and the deeply feature extraction is finished to realize the effective recognition for steel water status using wavelet package method. Experimental results indicate that the method has low realization cost and long operating life, requires little rebuilding for current devices, and its slag detection rate can reaches more than 95 percents.

引用

页码：141 / 146

页数：5

共 4 条

[1] Li P., Liu G., Tan D., The research of slag-detecting vibration method based on HMM in continuous casting, Proceedings of the ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, 4, pp. 229-234, (2005)
[2] Blok F.R., Dittmer R., Field distribution in rotationally symmetric moving conductors inside slag detection systems, Magnitnaya Gidrodinamika, 4, pp. 83-90, (1993)
[3] Wiskott L., Segmentation from motion: Combining gabor- and Mallat-wavelets to overcome the aperture and correspondence problems, Pattern Recognition, 32, 10, pp. 1751-1766, (1999)
[4] Yen G.Y., Lin K.C., Wavelet packet feature extraction for vibration monitoring, IEEE Conference on Control Applications-Proceedings, 2, pp. 1573-1578, (1999)

← 1 →