一种基于PSO-BP神经网络的重力数据插值方法  

A Gravity Data Interpolation Method Based on PSO-BP Neural Network

作　　者：谢心和 赵东明[1] 刘长青 XIE Xinhe;ZHAO Dongming;LIU Changqing(Institute of Geospatial Information,Information Engineering University,Zhengzhou 450001,China)

机构地区：[1]信息工程大学地理空间信息学院,河南郑州450001

出　　处：《测绘与空间地理信息》2023年第11期26-29,34,共5页Geomatics & Spatial Information Technology

基　　金：国家自然科学基金项目(42174008)资助。

摘　　要：重力数据在使用上需要通过进一步的精化和融合,并利用精确的插值技术进行加密处理,以满足实际应用的重力数据基础。本文提出一种基于粒子群算法(Particle swarm optimization,PSO)优化BP神经网络的方法,并应用于重力数据推估。通过仿真实验对比分析了PSO-BP神经网络、普通BP网络和传统克里金方法在重力异常插值上的效果,发现本文方法在均方差、差值稳定性上具有一定优势,但运算时间较长。利用澳大利亚中部的重力观测数据对PSO-BP神经网络进行训练,结果表明,本文方法在整个区域的插值效果上优于BP神经网络和克里金方法,补充高程数据作为PSO-BP神经网络的输入,能进一步提升利用该方法推估自由空气重力异常场的精度。Gravity data needs to be further refined and fused,and encrypted with accurate interpolation technology to satisfy the gravity data base of practical application.In this paper,a method based on Particle swarm optimization(PSO)to optimize BP neural network is proposed for gravity data estimation.The results of PSO-BP neural network,ordinary BP network and traditional Kriging method on gravity anomaly interpolation are compared and analyzed by simulation experiments.It is found that the proposed method has certain advantages in mean square error and difference stability,but it takes a long time to calculate.The PSO-BP neural network was trained using gravity observation data from central Australia.The results show that the interpolation effect of proposed method in this paper on the whole area is superior to the BP neural network and the Kriging method by adding elevation data as the input into the PSO-BP neural network to further improve the accuracy of estimation of the free air gravity anomaly field.

关 键 词：PSO-BP神经网络 克里金方法 重力异常 重力基准图 

分 类 号：P209[天文地球—测绘科学与技术]