测绘通报 ›› 2024, Vol. 0 ›› Issue (4): 48-53,82.doi: 10.13474/j.cnki.11-2246.2024.0409

• 学术研究 • 上一篇    

基于SBAS-InSAR与MA-PSO-BP的南京河西地区地表沉降监测及预测分析

毕凌宇, 孙承志, 乔申   

  1. 南京信息工程大学, 江苏 南京 210044
  • 收稿日期:2023-10-19 发布日期:2024-04-29
  • 通讯作者: 孙承志。E-mail:003453@nuist.edu.cn
  • 作者简介:毕凌宇(2000—),男,硕士生,研究方向为InSAR技术监测应用与预测分析。E-mail:2065883279@qq.com
  • 基金资助:
    高分专项航空观测系统“高分航空载荷自然资源调查应用示范项目”(04-H30G01-9001-20122)

Surface subsidence monitoring and predictive analysis in Hexi area of Nanjing based on SBAS-InSAR and MA-PSO-BP

BI Lingyu, SUN Chengzhi, QIAO Shen   

  1. School of Remote Sensing and Geomatics Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China
  • Received:2023-10-19 Published:2024-04-29

摘要: 针对南京河西地区城市化进展的不断加快及对该地区的沉降预测研究较少的问题,本文提出一种基于小基线集合成孔径雷达干涉测量(SBAS-InSAR)与滑动平均-粒子群优化-反向传播神经网络算法(MA-PSO-BP)的城市地表形变监测及预测模型。利用2020年3月—2022年3月的22景Sentinel-1A升降轨数据对南京河西地区进行沉降监测,获取研究区升降轨形变量,分析河西地区的沉降趋势与成因,并对监测得到的沉降值进行滑动平均插值,将其作为PSO-BP网络模型的样本输入,构建网络预测模型。结果表明,SBAS-InSAR技术能够有效监测城市长时间的沉降,南京河西地区存在不同程度的沉降,沉降速率为-25.3~20.5 mm/a。对比历史沉降研究,沉降趋势由北部向南部扩张,结合SBAS-InSAR沉降监测数据,分别与BP神经网络和PSO-BP神经网络预测模型进行对比,样本数据经过插值后沉降预测模型的精度最高。

关键词: 地表形变监测, 预测模型, 滑动平均插值, SBAS-InSAR, PSO-BP

Abstract: In view of the rapid urbanization in Hexi area of Nanjing and the few researches on settlement prediction in this area, this paper proposes a monitoring and prediction model of urban surface deformation based on small baseline subsets-interferometric synthetic aperture radar (SBAS-InSAR) and moving average-particle swarm optimization-backpropagation(MA-PSO-BP) neural network algorithm. The settlement monitoring of the Hexi area of Nanjing is carried out by using the 22 Sentinel-1A lift rail data from March 2020 to March 2022, the variable of the lifting rail in the study area is obtained, the trend and the causes of settlement in Hexi are analyzed, and the settlement value obtained is used as the sample input of the PSO-BP network model to construct a network prediction model. The results show that SBAS-InSAR technology can effectively monitor the long-term settlement of the city, there are different degrees of settlement in Hexi area of Nanjing, the settlement rate is -25.3~20.5 mm/a, compared with the historical settlement study, the settlement trend expands from north to south, combined with the settlement monitoring data of SBAS-InSAR, compared with BP neural network and PSO-BP neural network prediction model, the accuracy of the settlement prediction model after interpolation of sample data is the highest.

Key words: surface deformation monitoring, prediction model, moving average interpolatio, SBAS-InSAR, PSO-BP

中图分类号: