Application of ANN and SVM for Uncertainty Quantification and Propagation

摘要：

　　All measurements have error that obscures the true value.The error creates uncertainty about the quality of the measured value,which is requiring testing and calibration laboratories to provide estimates of uncertainty with their measurements.Measurement uncertainties include input uncertainty,the propagation of input uncertainty,the output uncertainty and the systematic error uncertainty.Several methods for estimating the uncertainty of measurements have been introduced for different kinds of uncertainty quantification,and two data mining methodologies-Artificial Neural Network (ANN) and Support Vector Machine (SVM) are used to build the unknown propagation model.This paper will discuss the quantification of measurement uncertainty (MU) and the separation of various uncertainty sources to MU and will discuss the advantages and limitations of SVM and ANN for building the propagation model of MU.

关键词： ANN SVM measurement uncertainty propagation quantification separation

作者: QinShu He ShiFu Xiao XinEn Liu

作者单位: Institute of Structural Mechanics, CAEP, Mianyang, China, 621900

会议类型: 国际会议

会议名称: the 2011 International Conference on Frontiers of Manufacturing Science and Measuring Technology (第一届制造科学与检测技术国际会议(ICFMM2011))

会议地点: 重庆

会议语种:英文

页码: 192-196

在线出版日期: 2011-06-23（万方平台首次上网日期，不代表论文的发表时间）