主成分分析在GPS高程拟合中的应用

摘要

为了合理确定多项式的阶数,克服参数间的相关性,提高GPS高程拟合的精度,提出采用主成分分析法,对相同的数据点分别采用常规多项式拟合与主成分分析拟合.苏通大桥首级施工控制网的GPS高程拟合计算结果表明,主成分分析法受参数个数与相关性影响小,计算结果稳定,而且能有效降低数据冗余度,是一种比较好的拟合方法.%In order to rationally determine the order of polynomial, to overcome the eorrelativity among parameters and to raise the fitting precision of GPS height, it was put forward that the conventional polynomial fitting and the principal component analysis were respectively employed for the same data point by means of the principal component analysis method. The calculated results of GPS height fitting for the first stage construction control network of Su-Tong Bridge indicate that the principal component analysis method is little affected by the number and the correlativity of parameters and has stable computational results. It can effectively reduce the data redundancy, thus it is a good method for fitting.