Research On Key Model Of GNSS Single-Frequency Single-Epoch Positioning For Bridge Monitoring

With the improvement of global navigation satellite system,GNSS precision positioning technology has gradually demonstrated its huge advantages in dynamic bridges deformation monitoring with its charateristics of real-time,all-weather and automation.It has been widely used in flexible highway bridges such as cable-stayed bridges and suspension bridges.The single-frequency single-epoch(SFSE)baseline solution uses double-difference processing technology,so that most of the errors in GNSS measurements can be well weakened.The single-frequency receiver has low cost,and the accuracy of multi/single-system double-difference solution is high enough,which make the SFSE baseline resolution suitable for bridge monitoring with the demand of high precision and dense monitoring points.However,it still has the following problems:(1)Due to the occlusion and multipath phenomenon in the bridge environment,the traditional stochastic model is difficult to describr the actual observation noise level;(2)the multipath error can not be eliminated by the double difference processing,seriously affecting the accuracy of the solution;(3)The ambiguity resolution success rate and monitoring accuracy of single GPS system is not high.In order to solve the above problems,the paper studies the stochastic model based on fixed solution residuals,multi-GNSS sideral filtering method based on KF-RTSS to mitigate multipath errors and multi-GNSS systems for bridge monitoring method with baseline length constraint.The main work and results of this paper can be summarized as follows:1)According to the characteristic that the observation noise remains basically unchanged for a short time,the weighting method based on the fixed solution residuals reflecting the accuracy of the observation values is studied,and the detailed solution flow chart and corresponding steps are designed.The experiment uses two sets of baseline data to compare different stochastic models of different systems.The results show that the weight method based on fixed solution residuals can effectively improve the success rate of ambiguity resolution,especially for GPS satellites with fewer average avalibale satellites.2)The operation laws of GPS,BDS and Galileo satellites are studied,and the relationship between GNSS multipath error and satellite repetition period is preliminarily established.The observation damain sidereal filtering method is used to mitigate multipath errors with the single-difference residual reconstruction technique.Based on this,the kalman filter and RTS smoother is introduced to denoise single-difference residual and extract the multipath error model.Wavelet filter and EMD decomposition are also used as comparison.The extracted multipath error model is corrected,and the ENU direction deviation of the baseline solution is greatly improved.3)For the single GPS baseline resolution,the accuracy of the floating point solution is not high,resulting in poor ambiguity success rate.The GPS/Galileo single frequency single epoch technology is used to monitor the bridge displacement combined with the baseline length constraint.The performance of single GPS,baseline length constrained GPS,GPS/Galileo and baseline length constrained GPS/Galileo was compared using two sets of GPS/Galileo monitoring data.The results show that the introduction of Galileo and baseline length constraints can effectively improve the ambiguity success rate of GPS for bridge monitoring.