Design of GNSS-RTK Landslide Monitoring System Based on Improved Raida Criterion

TL;DR

This paper presents a GNSS-RTK landslide monitoring system that integrates improved error detection and noise filtering techniques with IoT and edge computing, achieving high accuracy and reliability in real-time deformation monitoring.

Contribution

It introduces an enhanced Raida criterion-based error detection method combined with filtering and edge computing to improve GNSS-RTK landslide monitoring accuracy and efficiency.

Findings

Real-time deformation monitoring accuracy of 10 mm achieved

System's data transmission is reliable and efficient

Monitoring error is smaller than single-GPS system

Abstract

Aiming at the problem that GNSS-RTK technology cannot effectively monitor landslides due to gross errors and high-frequency noise during landslide monitoring, a GNSS-RTK landslide monitoring system based on the improved Raida criterion(3{\sigma}) was designed. The system uses Raspberry Pi as the control core, GNSS-RTK technology to complete deformation data collection, and combines NB-IoT to achieve data transmission and cloud storage. To further improve the monitoring accuracy, real-time gross error detection and high-frequency noise removal method based on the improved Raida criterion(3{\sigma}) and Butterworth low-pass filtering is proposed, combined with edge computing devices to complete real-time data processing, and reduce the pressure of cloud computing. The experimental results show that the system's data transmission is reliable and efficient. After gross error elimination and noise processing, real-time deformation monitoring of 10 mm in the horizontal and vertical directions is achieved; simultaneously, the monitoring error of the GNSS combined system is smaller than that of the Single-GPS system.