De-nosing root tensile test data of Pinus tabulaeformis
WANG Ping-Hua, Chen-Li-Hua, Wang-Hua-Shun, Zhou-Shuo, Ji-Xiao-Dong, Diao-Gong-Hua
1.College of Soil and Water Conservation, Beijing Forestry University, Key Lab. of Soil & Water Conservation and Desertification Combating, Ministry of Education;2.College of Science, Beijing Forestry University: 100083, Beijing, China
Abstract:The database of single root tensile test is often confused by a series of noise because of test environment, testing machine random error, artificial operation, which greatly influences accurate analytical result. In order to improve analytical precision, the data is needed to be pretreated. The purpose of the study is to eliminate noise of root tensile test data of Pinus tabulaeformis with gauge length 250mm by wavelet analysis. Several multi-resolution wavelet de-nosing methods are applied, including from 3 to 31 point smoothing method, hard threshold method, fixed threshold method, mini-max threshold method and rigorous sure threshold method. Results show that among various window width of smoothing methods, 9 point smoothing method is proved to have more marked de-noising effect. In addition, the wavelet transform has particular advantage on noise elimination of the database of single root tensile test while reserving the useful information of data, thus improving the quality of test data and analytical accuracy.