GNSS Deformation Information Recognition And Early Warning Based On Control Chart Method

Global Navigation Satellite System(GNSS)is a space-based radio positioning System that can continuously provide users with time information,THREE-DIMENSIONAL coordinates and speed.GNSS technology is applied in every aspect of people's life,which greatly promotes the birth and innovation of many technologies.GNSS,with advantages of all-weather,real-time,rapid and high precision,has gradually become the main monitoring means for deformation monitoring at home and abroad.GNSS precise positioning technology is an effective supplement and verification means for traditional external deformation monitoring.The integration of GNSS technology with computer technology,data communication technology and data processing and analysis technology can realize the automation degree of data acquisition,transmission,management,deformation analysis and prediction.Therefore,the identification and early warning of deformation information in monitoring data is of great significance to the safety of people's life and property.In this paper,using the method of control chart in the time series of GNSS deformation information identification and early warning,first of all to do the CUSUM control chart and Shewhart control chart recognition performance comparison and analysis,combining with the characteristics of both considering joint control method was adopted to realize to warn the equilibrium of the deformation size deviation,then use a simple structure,easy calculation of parameter adaptive accumulation and control chart to realize equilibrium monitoring,to improve the computing efficiency and universality.Finally,aiming at the problem of high error warning rate of large deviation caused by accumulation and algorithm,a comparative analysis on the parameter selection and detection performance of the moving weighted average algorithm is proposed,and it is concluded that the algorithm has strong universality and ability to identify deformation interval.The main research contents of this paper are as follows:(1)Using the control chart in the statistical process control method to conduct abnormal inspection on the GPS monitoring sequence of large buildings is of great significance for early detection of anomalies and reduction of losses.This paper add different migration in simulated data to compare CUSUM control chart and Shewhart control chart of detection capability,the results showed:Shewhart control chart for more than 2 times the standard deviation of the migration have good detection ability,control figure has to 1 times the standard deviation of the offset detection ability,but can produce false positives,with the increase of standard deviation cannot offset the locations of the visual performance.According to the characteristics of CUSUM control chart with good detection ability of small offset and Shewhart with good detection ability of large offset,a joint control chart of CUSUM and Shewhart is proposed.Under certain conditions,its detection effect is better than that of using one kind of control chart alone for process control.(2)In this paper,GNSS time series is processed by combining rank statistics with adaptive CUSUM control chart.Simulation experiments show that NAC control chart has good identification and warning ability for size offset by adjusting 'the size of parameters.Compared with the CUSUM control chart,the false alarm rate of the two control charts is lower for unknown offset,and the performance of the two control charts from deformation occurrence to early warning signal is similar.Considering that the NAC control chart is more concise in calculation,it is considered that the NAC control chart has better applicability.(3)Due to its own algorithm,CUSUM control graph is unable to identify the true interval of deformation due to large offset.Moreover,its algorithm is established on the basis of sequential probabity-ratio test,and it has limitations on the distribution of samples.There will be many limitations in the application,and the method of data replacement will reduce the detection efficiency.In this chapter,it is proposed to use EWMA algorithm to establish deformation warning model.EWMA algorithm is not sensitive to sample distribution and can control non-normal data,which has a good detection effect on deformation information in monitoring data.EWMA control chart has better detection results than classical CUSUM control chart under certain conditions.Figure[26]table[6]reference[76]...