Kalman Prediction Of Surface Dynamic Deformation Based On High Frequency GNSS Coordinate Sequences

With the accelerating process of urbanization,the required economic construction has continued to increase,and construction facilities such as high-rise buildings,large-scale water conservancy projects,gob areas in mining areas,and bridge projects have continued to increase.This makes dynamic deformation monitoring of the ground surface an important part of modem research.Dynamic deformation has an extremely complex internal law,which is not caused by the interference of a single factor.Each factor contributes.Grasping the degree of deformation of deformation body is the primary purpose of deformation monitoring.According to the scope of deformation degree,it is judged whether it is within the safety system.If it is beyond the safety scope,we must take measures to prevent disasters.At this point,we can see that the purpose of deformation monitoring is not only to monitor the deformation,it is more important to prevent disasters and ensure personal safety.This article focuses on the village of Qianwei,which is located at the working face of 1222(1)in Zhujidong Coal Mine,and monitors the villages to ensure the safety of the village.First,by establishing a real-time automated monitoring system,high-frequency GNSS data is acquired with a time interval of 1 s.Secondly,kalman filtering method is used to process and analyze the acquired high-frequency GNSS data.Finally,through the establishment of an effective forecasting and early warning model,scientific disaster prevention and control is carried out.The article chooses two kinds of filtering methods for experiment and analysis,which are standard kalman filtering algorithm and variance compensation adaptive kalman filtering algorithm.The experimental comparison shows that the variance-compensated adaptive kalman filter algorithm can better avoid the filter divergence.Compared with the standard kalman filter algorithm,the precision is higher and the forecast value is closer to the true value.At the same time,in order to determine the time interval for the best conventional measurement,the experiments were conducted at intervals of 1 day,5 days,10 days,and 15 days,respectively.Through comparison experiments on observed data at different time scales,it was found that the time interval was different.The shorter and the higher the precision,the better the accuracy of the fourth-class level can be achieved.However,the shorter the time interval,the higher the required cost and the greater the difficulty of operation.Considering the above two points,the paper finally determines the time interval of 10 days as the conventional measurement.time interval.The establishment of real-time automated monitoring system and kalman forecasting model can more accurately and timely find the deformation and trend of local surface,and further improve the safety of surface dynamic deformation monitoring.