Research On BDS Dual-antenna Key Technology Of High Precision Deformation Monitoring And Its Application

At present,BeiDou Navigation Satellite System(BDS)can provide navigation,timing and short message information service in the Asia-pacific region.The application of BDS positioning technology in the field of deformation monitoring has become one of the research hotspots.Accuracy,cost,and real-time are the key issues to be resolved in BDS deformation monitoring,which normally adopts the double-differential model to carry out static position,meanwhile the calculation cycle is generally several hours,and the hardware is the multi-frequency multi-model geodetic receiver(>? 46,000).Although the static position can realize the position with millimeter accuracy,it still does not meet the cost and real-time requirement of the majority structure monitoring,and it is urgent to research BDS single epoch high-precision positioning algorithm based on the low cost receivers.The research was based on the National Natural Science Foundation of China(51378057)and the Beijing Key Laboratory of Urban Spatial Information Engineering(2016101).By systematically summarizing the hardware devices of Global Navigation Satellite System(GNSS)at home and abroad,the algorithm of ambiguity fixing and the model of high-precision positioning,and taking clock-synchronized dual-antenna single-frequency receivers as hardware device,the influence of BDS single-difference precision stochastic model on single-frequency ambiguity fixing was studied.Then based on it,an improved par ambiguity algorithm with deformation characters and maximum displacement constraints was proposed.Through outdoor experiments,we researched the main error characteristics of ultra-short baseline and short baseline coordinate time series and the methods of error post-processing elimination,moreover UPD(Uncalibrated Phase Delay)real-time corrected filtering model and BDS/GPS displacement detection model using robust weighting based on moving window were built.Combining with the monitoring examples of engineering structures,the effectiveness of BDS dual-antenna deformation monitoring technology was verified.The main contents of this paper are as follows:(1)BDS single-difference precise function and stochastic models were constructed,and it was proved that the elevation accuracy of the single-difference model(UPD = 0)was obviously higher than that of the double-difference model with theoretical derivation and practical data based on clock-synchronized dual-antenna receivers.The significance of the correlation between the B1 frequency pseudorange and carrier observations was proved based on statistical hypothesis testing,indicating the necessity of constructing the precision stochastic model.(2)The influence of precision and empirical stochastic models on BDS single frequency ambiguity fixing was studied.The observation and calculation model considering the correlation between observations and the additional baseline length constraint was built,then combined with the static experimental data,the ambiguity fixed success rate was improved by using the precision stochastic model,and the critical value of Ratio was set to 3,which could effectively avoid the generation of abnormal ambiguity.The multipath error was not considered in the precision stochastic model,and the ambiguity fixing effect of the short baseline was slightly lower.(3)According to the poor effect of BDS single frequency ambiguity fixing,the algorithm of IParLAMBLDC(Improved Partial LAMBDA with Baseline Length and Deformation Characteristics Constraints)was proposed,improving the ambiguity searching efficiency and fixing success rate remarkably,slightly lower than those of GPS LI frequency.Through numerical simulation of frozen soil roadbed of Gonghe-Yushu highway,the Lijiaxia dam in Qinghai and the debris slope of 330 kV power transmission channel in Hui County,Gansu Province,and analyzing geometric relation between the slope main sliding direction and SCS(Site-Centric coordinate system),the method of determining the deformation characteristics and maximum displacement of the structure was defined.(4)The major characteristics of the time series residual error of ultra-short baseline and short baseline were analyzed.Combined with outdoor static test,it was found that the positioning accuracy of single-difference model for ultra-short baseline was mainly affected by UPD,and the accuracy of the single-difference model with UPD uncorrected was basically the same as that of double-difference model.The accuracy of single-difference model with UPD corrected was somewhat improved,especially the increasing amplitude is the maximum in the elevation.The positioning accuracy of short baseline was mainly affected by the multipath effect.After the multipath error was eliminated,the positioning accuracy had been greatly improved.(5)Because the UPD moving average filtering model could not meet the real-time requirements of structural monitoring,UPD ARMA(Auto-Regressive and Moving Average)real-time correction filtering model was proposed.Through the static and displacement loading tests,FISHER hypothesis testing and displacement detection consistency were used to prove the superiority of UPD ARMA real-time correction filter model in displacement detection.Combined with the monitoring example of ventiduct frozen soil roadbed,the effectiveness of this model for deformation monitoring of frozen soil roadbed was validated.(6)Based on the poor redundancy and positioning accuracy of BDS single-frequency model,this paper proposed BDS/GPS Kalman filter model with robust weighting based on moving window for short baseline(1km).Combined with outdoor tests,it was found that the equivalent weight anti-error could speed up the outline coordinates to the accurate coordinates and improve the coordinate accuracy.The displacement detection accuracy would increase significantly with the increasing of the window.The detectable displacement convergence rate was inversely proportional to the moving window parameter.The algorithm is suitable for high-precision monitoring of structures with creep features.(7)BeiDou deformation monitoring pre-warning and forcast analysis system was designed.Based on C#network technology and using PostgreSQL as the data management platform,BeiDou deformation monitoring pre-warning and forcast analysis system was developed,which integrated such subsystems as data acquisition,data transmission,data processing and analysis,data management,data release and etc.Taking the Lijiaxia dam as an example,BDS/GPS monitoring data query,real-time display,report generation,predictive analysis,and early-warning functions were realized,and the advantages of clock-synchronized receiver technology for deformation monitoring were analyzed.(8)As higher cost was the main limiting factor in the application of geodetic receivers in dam deformation monitoring,an improved GNSS monitoring scheme was proposed.Three-dimensional finite element model was built to simulate the main deformation part of the Lijiaxia dam,which was selected as GNSS monitoring station,and only one clock-synchronized dual-antenna receiver was set to evaluate the dam operating status,greatly reducing the monitoring cost.The dam displacement warning standard was determined based on the relationship between the corresponding surface displacement and safety factor,then the effectiveness of this standard was verified with fuzzy neural network.(9)Taking the example of landslide geological disaster monitoring of 330kV power transmission channel in Hui County,Gansu Province,BDS/GPS landslide monitoring technological scheme suitable for a wide range of linear engineering structures was proposed.Based on the AHP(Analytic Hierarchy Process)method,the stability classification of the landslide was evaluated with rough neural network.The slope selected belong to hazard grade region was spec1ally monitored,improving the use efficiency of GNSS receiver equipment resources.The corresponding numerical analysis model was established to simulate the landslide area of the debris slope,and it was verified that BDS/GPS model with robust weighting based on moving window could be used in monitoring the landslipe with creep characteristics.