基于改进的二次曲面模型在GPS高程异常拟合中的应用研究

Research on Application of Improved Quadric Surface Model in GPS Height Anomaly Fitting

GPS高程拟合的方法主要有二次曲面拟合、多面函数拟合、最小二乘配置拟合、BP神经网络拟合等,但是这些方法在应用上都存在一定的局限性,特别是在地形较为复杂的区域会存在较大的精度损失。基于这些问题,本文提出了一种改进的二次曲面拟合模型进行高程拟合。该模型的基本思路是将二次曲面模型和最小二乘配置拟合模型进行组合应用,首先利用二次曲面拟合模型计算各拟合点的残差序列;然后利用最小二乘配置模型获取残差序列模型;最后对各点的高程异常值进行残差改正,得到改正后的高程异常,并对该方法的结果进行精度评价分析,验证了方法的可行性,对以后的工程实践有一定的借鉴意义。

GPS elevation fitting methods mainly include quadric surface fitting,polyhedral function fitting,least square configuration fitting,BP neural network fitting,etc.However,these methods have certain limitations in application,especially in areas with more complex terrain will have a greater loss of accuracy.Based on these problems,this paper proposes an improved quadric surface fitting model for elevation fitting.The basic idea of the model is to combine the quadric surface model and the least squares configuration fitting model.First,use the quadric surface fitting model to calculate the residual sequence of each fitting point,and then use the least squares configuration model to obtain the residual sequence model.Finally,the residual correction is performed on the elevation anomaly value of each point,and the corrected elevation anomaly is obtained,and the accuracy evaluation and analysis of the results of the method are performed,which verifies the feasibility of the method and has certain reference significance for future engineering practice.