摘要
管道漏磁检测技术利用漏磁原理对管道进行无损检测.传统的人工检测方法通常使用漏磁检测器采集的管道漏磁数据,绘制出漏磁信号曲线,然后根据曲线的变化特性对管道上的缺陷和组件进行人工判别,这种方法效率低下且具有很强的主观性.随着人工智能技术的快速发展,许多基于人工智能的漏磁检测方法被提出,可实现更加高效和更加准确的智能检测.本文对管道漏磁检测的智能方法进行了综述,首先简要介绍了漏磁检测的基本原理和漏磁检测器的组成结构,随后重点阐述了管道漏磁检测中的机器学习方法(含基于分类的方法、基于目标检测的方法和多分量方法)、基于知识的智能专家系统和多传感器融合方法,最后进行了总结,并讨论了当前智能方法仍然存在的问题.
The principle of magnetic flux leakage(MFL)is used in the pipeline MFL detection technology to carry out non-destructive detection of pipelines.The traditional manual detection method usually draws the MFL signal curve by using the pipeline MFL data collected by the MFL detector,and then manually identifies the defects and components on the pipelines according to the change characteristics of the curve.This method is inefficient and subjective.With the rapid development of artificial intelligence technology,many MFL detection methods based on artificial intelligence have been proposed,which can achieve efficient and high-precision intelligent detection.In this paper,the intelligent methods have been reviewed for pipeline MFL detection.Firstly,the basic principle of MFL detection and the structure of MFL detector have been briefly introduced.Then,the machine learning methods(including classification based method,object detection based method and multi component method),knowledge based intelligent expert system,and multi-sensor fusion methods in pipeline MFL detection have been surveyed.Finally,this paper has been summarized,and the problems of current intelligent methods have been discussed.
作者
杜文飞
李春光
万四海
DU Wenfei;LI Chunguang;WAN Sihai(Polytechnic Institute, Zhejiang University, Hangzhou 310015, China;Ningbo Research Institute, Zhejiang University, Ningbo Zhejiang 315100, China;College of Information Science and Electronic Engineering, Zhejiang University, Hangzhou 310027, China;China Wuzhou Holdings Limited, Xuchang Henan 461200, China)
出处
《西南师范大学学报（自然科学版）》
CAS
2022年第6期1-7,共7页
Journal of Southwest China Normal University(Natural Science Edition)
基金
国家自然科学基金项目(61631003).
关键词
管道缺陷
漏磁检测
人工智能
异常检测
pipeline defects
magnetic flux leakage detection
artificial intelligence
anomaly detection
