摘要
通过对一种套损设备新型双远场电磁聚焦测厚仪的实验数据算法处理,获得套管厚度变化的检测算法.在数据处理中,包络线法能够快速、准确地提取测得波形的最大幅值,避免了因在线数据的出现干扰造成的误判,提高了检测算法的抗干扰能力;在BP网络训练改进措施中,影响因子数据变化率极大地改善了BP网络的稳定性,提高了网络的通用性,继而提高了检测算法的精度.实验结果表明,此算法能够准确检测到套管因破损变化的厚度,为套管破损的进一步定量化提供了保证.
After a large number of experiments ,a detection algorithm is studied which can get the casing thickness variation by an algorithm processing for the experimental data of this casing damage equip -ment .In data processing ,the envelope method presented in the paper can quickly and accurately pick out the maximum amplitude of the measured waveform and avoid the misjudgment caused by interference of online data emergence .Therefore ,it greatly improves the anti-interference ability of the detection al-gorithm .In the BP network training improvement measures ,data impact factor --- data change rate is added which can greatly improve the stability of the BP network ,improve the universality of the net-work ,and then improve the accuracy of detection algorithms .The experimental results show that the improved detection algorithm can accurately measure and get the varying thickness caused by the damage of the casing .It provides a guarantee for further quantitative of casing damaged .
出处
《西安工程大学学报》
CAS
2014年第1期84-88,共5页
Journal of Xi’an Polytechnic University
关键词
套管厚度变化
BP算法
包络线法
数据变化率
casing thickness variation
BP neural network
envelope method
data change rate