摘要
根据广义最小二乘法原理,用直线或曲线去拟合两组变量.提出的新方法;视两组观测值均为具有误差的变量.更为切合实际.所求出的参数估值为无偏估值.
In this paper general least squares methad is used to fit strait line or curve and two sets of observations are taken as variables. In the past the empiric formula was found with conventional method which assumed that in two sets of observations one set were constants and another set were variables. The author presents a new method that treats both of two sets of observations as variables.This method is geared to reality and the estimates of unknown parameters found with this method are unbiassed estimates.
关键词
工程测量
平差
经验方程
直线拟合
empiric formula
general leas t squares method
strait line fitting
curve fitting
adjust of indiret observations
matrix
