随机-模糊线性回归模型的参数估计及应用

Regression Parameters Estimation of Random-Fuzzy Linear Regression Model and its Application

导出了随机-模糊线性回归模型参数的估计量,证明了参数的估计量为无偏估计,同时推导了参数估计量数字特征和回归方程相关系数的计算公式。将该模型应用于岩石样本抗剪强度实验数据处理中,通过与传统的随机一元线性回归对比分析,表明使用该方法得到的力学参数更具代表性。

The regression parameters estimation exprssions of a random-fuzzy linear regression model are deduced. It is proved that they are an unbiased estimator. The formulae for calculating the numerical characteristics of regression parameters estimation and the correlation coefficients of the regression equation are derived. Based on the experimental data of triaxial compression tests, the calculated results are more realistic and reasonable compared with the linear regression method.