Research On Multipath Effect Of GPS In Structural Health Monitoring

Because the technology of GPS satellite positioning has many advantages, such as its high positioning accuracy, all-weather work, easy of operation and so on, it has been successfully applied in structural health monitoring domain of large-scale civil engineering. However, the GPS satellite signals are easy to be influenced by many kinds of disturbance errors, such as ionospheric and tropospheric errors, orbital errors, clock errors and multipath effect and so on. In structural monitoring, when the distance between reference station and rover station is short, its public errors can be eliminated through difference GPS. But the monitoring station's relevance of multipath error is very weak, even if the baseline is very short, multipath's influence is also very different between two stations. so the multipath effect is difficult to be removed, and it has become the biggest restrictive factor of high accuracy positioning. According to above problems, the main aspects of theoretical and experimental research are as follows:(1) Theoretical research for GPS multipath error has been carried on, and numerical simulation for GPS multipath error of common GPS carrier phase survey has been carried on. The results indicates that the satellite signal's angle of arrival, the carrier wave length, the reflection factors of reflecting objects and the horizontal distance between receiver antenna and the reflecting surface are primary factors that affect the multipath error. Moreover the multipath error increases along with reflection factors increase. The multipath error reduces with the horizontal distances increase, and the reduction gets smaller.(2) The insufficiencies are systematically summarized of existing GPS data post-processing methods, and particle filtering algorithm has been studied in GPS multipath denoising, and the improved particle filtering algorithm is applied in the numerical example. The results prove that the improved particle filtering algorithm can effectively eliminate the Gauss noise in GPS multipath siganls.(3) Merits and drawbacks of different transmission modes are compared, on the basis of it, and five commonly used arrangement methods of GPS monitoring system are designed. According to actual situation, a set of simulating and monitoring system for GPS multipath signals was established. Validity is checked for the system by the flexible steel frame model, and spectral analysis results indicate carrying on the monitor using the system is effective. (4) According to dielectric constant variation range of mediums, dielectric constant test experiment was first carried on to the medium of representative building materials, and five materials were finally chosen as the reflecting mediums of the multipath error experiment based on the experiment results.(5) In GPS multipath error experiment, continuous real-time dynamic observation data was obtained by means of placing reflecting objects at different heights and distances from the receiver antenna. The improved particle filtering algorithm is being used in the GPS data post-processing, and it effectively eliminates the system noise of GPS, thus the influence which the multipath error causes can be obtained. By comparison, it is found that the multipath errors which caused by the same reflecting medium at different distances and the different reflecting mediums at same distance have certain regularity, and which coincides with the numerical simulation results of multipath error very well. It provides the basis for the further establishing the revision model and weakening the multipath error.