基于超声辅助的市政地下管线探测方法  被引量:1

Detection method of municipal underground pipelines based on ultrasonic assistance

在线阅读下载全文

作  者:艾朋 Ai Peng(Sihui City Multi Regulatatory Compliance Guidance Center,Sihui 526200,China)

机构地区:[1]四会市多规合一指导中心,广东四会526200

出  处:《工程勘察》2023年第11期63-68,共6页Geotechnical Investigation & Surveying

摘  要:为实现市政地下管线的全面管理,及时检测管线缺陷,本文提出基于超声辅助的市政地下管线探测方法。根据超声波发射后的传播和反射波形结果及其穿透地下管线时的能量变化结果,获取地下管线缺陷信号;采用变分模态分解方法,分解处理获取的缺陷信号,得到管线缺陷特征信息后,将其输入优化的随机森林模型中分类,对分类结果进行投票,取投票数量最多的决策结果为地下管线的最终探测结果。结果表明,地下管道信号提取效果较好,可分性测度指标均在0.92以上,最高值为0.98;汉明损失指标结果均低于0.05,检测性能良好;能够可靠地完成地下管线不同类别缺陷的分类检测,为地下管线的全面管理提供可靠依据。In order to realize the comprehensive management of municipal underground pipelines and timely detect pipeline defects,this paper proposes a detection method of municipal underground pipelines based on ultrasonic assistance.According to the propagation and reflection of ultrasonic wave and the change of energy when it penetrates underground pipeline,the defect signal of underground pipeline is obtained.The obtained defect signals are decomposed and processed by the variational mode decomposition method.After the information of pipeline defect characteristics is obtained,it is input into the optimized random forest model for classification.The classification results are voted,and the decision result with the largest number of votes is taken as the final detection result.Results show that the extraction effect of underground pipeline signals by this method is good,and the separability measure indexes are all above 0.92,with the highest value of 0.98.The results of Hamming loss index are all lower than 0.05,indicating good detection performance.It can reliably complete the classification and detection of different kinds of underground pipeline defects,and provide a reliable basis for the overall management of underground pipelines.

关 键 词:超声辅助 地下管线探测 管线缺陷信号 反射波形 

分 类 号:TU990.3[建筑科学—市政工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象