학술논문
基于优化小波分量的高铁桥墩沉降异常探测方法 / Outlier detection method of bridge pier settlement in high-speed railway based on optimized wavelet component
Document Type
Academic Journal
Author
支君豪; 龚循强; 杨忠; 鲁铁定; 汪宏宇; 罗升; ZHI Junhao; GONG Xunqiang; YANG Zhong; LU Tieding; WANG Hongyu; LUO Sheng
Source
测绘工程 / Engineering of Surveying and Mapping. 33(1):41-46
Subject
Language
Chinese
ISSN
1006-7949
Abstract
针对高铁桥墩沉降数据中传统异常值探测方法存在探测结果不理想的情况,提出一种基于优化小波分量的高铁桥墩沉降异常值探测方法.所提方法在拉依达准则法的基础上,通过优化小波分量降低小波分量中的噪声影响并充分利用小波分量中的有效信息.利用模拟实验数据和工程实例数据进行实验,并与拉依达准则法、IQR法、MAD法以及4种传统小波组合方法进行对比分析.实验结果表明,文中所提方法相较于单一异常值探测方法和其它小波组合方法能够探测出更多的异常值,且探测结果更加可靠.
In view of the unsatisfactory detection results of the traditional outlier detection method of bridge pier settlement in high-speed railway,an outlier detection method of bridge pier settlement in high-speed railway based on the optimized wavelet components is proposed.The proposed method is based on the pauta criterion by optimizing the wavelet components,so as to reduce the effect of noise in the wavelet components and make full use of the effective information in the wavelet components.Experiments are conducted using simulated experimental data and engineering example data to compare and analyze with the pauta criterion method,IQR method,MAD method,and four traditional wavelet combination methods.Experimental results show that the proposed method can detect more outliers and is more reliable than the single outlier detection method and other wavelet combination methods.
In view of the unsatisfactory detection results of the traditional outlier detection method of bridge pier settlement in high-speed railway,an outlier detection method of bridge pier settlement in high-speed railway based on the optimized wavelet components is proposed.The proposed method is based on the pauta criterion by optimizing the wavelet components,so as to reduce the effect of noise in the wavelet components and make full use of the effective information in the wavelet components.Experiments are conducted using simulated experimental data and engineering example data to compare and analyze with the pauta criterion method,IQR method,MAD method,and four traditional wavelet combination methods.Experimental results show that the proposed method can detect more outliers and is more reliable than the single outlier detection method and other wavelet combination methods.