根据拟合推估两步极小解法的基本原理,将随机场分为趋势性和随机性两部分,对趋势性部分采用函数模型拟合后,用剩余残差拟合随机性信号协方差函数,再求解趋势性模型参数和随机信号。探讨了该方法在地图坐标变换中的应用,结果表明,基于两步极小的拟合推估新解法的结果明显优于最小二乘的结果。
After having reviewed the basic collocation theory, the principle of the collocation by two minimization steps is introduced. Considering that the stochastic signal contains not only the stochastic information but also some systematic trend, after the systematic trend parameter having been calculated and then the residual was used to compute the covariance function of the stochastic information. With the refitted covariance function, the stochastic signal was evaluated and the systematic trend parameters were estimated again. An actual digital map case shows that the solution accuracy with this method for collocation is better than that with the affine transformation by least squares.