摘要
变形监测中获得的观测数据不可避免地存在着噪声,如何从受噪声干扰的数据序列中提取特征信息,提高变形监测的精度是变形监测系统所涉及的关键技术问题之一。在赵瑞珍的阈值算法基础上,提出了一种改进的阈值算法,进而给出改进的模平方处理法。理论分析、工程实例的结果表明,改进的模平方处理法能够更有效地克服软阈值方法恒定偏差的缺点,进一步提高了消噪的精度和可靠性。
Noise in observations of deformation monitoring is inevitably. How to extract diagnostic information from data and improve deformation monitoring precision is a critical problem in the deformation monitoring system. Based on the threshold algorithm by Rui-Zhen Zhao, an improved threshold algorithm is presented. Furthermore, an improved modulus square processing method is proposed. The results of theoretical analysis and engineering projects show that the improved modulus square processing method can be more effective to overcome the shortcomings of constant bias in soft threshold method, and further improves the precision and reliability.
出处
《工程勘察》
CSCD
2012年第1期53-56,共4页
Geotechnical Investigation & Surveying
基金
国家自然科学基金项目(41071294)
广西自然科学基金资助项目(0640178)
广西科学基金资助项目(0991023)
广西研究生科研创新计划项目(2010105960816M37)
关键词
变形监测
模平方处理
阈值
信噪比
deformation monitoring
modulus square processing
threshold value
SNR
