Abstract:Due to the influence of GNSS hardware equipment, communication link, and observation environment, GNSS displacement monitoring data often contains gross errors, which cannot reflect the real deformation characteristics. In order to solve this problem, we propose to apply the robust random cut forest(RRCF) algorithm to real-time GNSS displacement gross error detection. The simulation data processing results show that the accuracy, precision and recall rate of real-time gross error anomaly detection of RRCF algorithm are better than 95%, 98% and 96%, respectively. The results of geohazard displacement monitoring data show that when outliers occur in GNSS displacement monitoring data, the detection results of RRCF method are in good agreement with the actual outliers, and have a low misjudgment rate. Overall, the RRCF algorithm can provide relatively high accuracy and availability for real-time outlier detection of GNSS displacement monitoring data.