一种基于灰色预测理论和抗差自适应Kalman滤波的滑坡监测算法

针对当前的山体滑坡监测技术监测精度低、实时性差、自动化程度低的问题，提出了一种基于灰色预测理论和抗差自适应Kalman滤波的滑坡监测技术。该技术使用抗差自适应Kalman滤波技术,对包括实时动态(RTK）载波相位差分定位数据、无人机摄影测量数据、土工带传感器数据在内的多源数据进行融合分析，将滑坡形变监测精度提高到了mm级。RTK技术和土工带传感器的使用克服了天气状况、植被覆盖对滑坡监测的影响。使用灰色预测理论对山体滑坡监测点进行形变预测，结合蠕变切线角判据，该技术实现了对山体滑坡预警等级的划分。仿真实验结果显示，该山体滑坡监测技术能够成功实现山体滑坡预测预警功能。

A Landslide Monitoring Algorithm Based on Grey Prediction Theory and Robust Adaptive Kalman Filter

Aiming at the problems of low monitoring accuracy, poor real-time and low automation of current landslide monitoring technology, a landslide monitoring technology based on grey prediction theory and robust adaptive Kalman filtering is proposed. Robust adaptive Kalman filtering technology is applied to fuse and analyze multi-source data, including real-time dynamic (RTK) carrier phase differential positioning data, UAV photogrammetric data, and geobelt sensor data, so as to improve the accuracy of landslide deformation monitoring to the millimeter level. RTK technology and geobelt sensor are applied to overcome the influence of weather conditions and vegetation coverage on landslide monitoring. The grey prediction theory is applied to predict the deformation of the monitoring points of the mountain landslide. Combined with the criterion of the creep tangent angle, the classification of the early warning grade of the mountain landslide is realized. The simulation results show that the landslide monitoring technology can successfully realize the landslide prediction and early warning function.