信息熵在缺陷漏磁信号量化中的应用

基金项目:

全国高校博士点青年教师基金资助项目(20092322120001)；黑龙江省自然科学基金资助项目(A200907)；黑龙江省自然科学基金资助项目(E201016)

详细信息

The Application of Information Entropy in Quantization of Magnetic Flux Leakage Signals of Defect

摘要: 缺陷漏磁信号的量化问题是漏磁信号研究领域中的一个重要方向。对缺陷漏磁信号进行FFT变换，根据变换后缺陷信号的幅值谱序列，对漏磁信号的信息熵进行了特征提取。利用提取的信号特征量，结合BP网络，完成了缺陷信号长度、深度的量化，量化精度分别达到了88.48%和84.17%。理论分析与试验研究证明了信息熵应用于漏磁信号量化过程中的可行性。
- 漏磁检测 /
- 信息熵 /
- 量化 /
- BP网络
Abstract: The quantization of defect magnetic flux leakage signal is always an important direction in the field of magnetic flux leakage signal processing. In this paper, Fourier transform was applied to magnetic flux leakage signals. By the use of the amplitude spectral sequence of defect magnetic flux leakage signal, information entropy was extracted. The length and depth of defect were quantified through the combination of signal characteristic and BP network, and the quantization precision was estimated reach 88.48% and 84.17%. It was proved that information entropy was feasible in the quantization of defect magnetic flux leakage signal by theoretical analysis and experimental study.
- Magnetic flux leakage testing /
- Information entropy /
- Quantization /
- BP network

[1]	李鹤龄.信息熵、玻尔兹曼熵以及克劳修斯熵之间的关系——兼论玻尔兹曼熵和克劳修斯熵是否等价［J］.大学物理,2004,23(12):37-40.
[2]	王太勇，刘兴荣，秦旭达，等.谱熵分析方法在漏磁信号特征提取中的应用［J］.天津大学学报,2004,37(3):216-220.
[3]	刘红星，左洪福，姜澄宇，等.信号频谱的二维向量及其应用［J］.中国机械工程,1999,10(5):537-539.