Research And Application Analysis Of BDS/GNSS Bridge Deformation Monitoring Algorithm In Complex Environment

Bridge is important for the development of society and economy,so timely and effective health monitoring is essential for it.GNSS has become an important sensor for bridge deformation monitoring,because of its all-weather,high-precision,real-time and full-automation advantages.The time series and frequency series could be obtained through the data processing,which can provide a variety of data reference for bridge health monitoring.The high-precision solution is the basis of the time series and frequency series analysis.The monitoring points should be located at the places which reflect the bridge’s characteristics.The occlusion of the buildings and passing vehicles,environmental errors and measurement noise will cause uncorrect ambiguities and low precision.Aimed at that,the main content and achievements are as follows:(1)The manuscript summarizes the development、theory and method of GNSS bridge deformation monitoring.On this basis,the advantages and the remaining problems of the prevenient algorithms are stated in details.(2)The influence of noise on the monitoring accuracy of GNSS bridges is analyzed from the results of data quality inspection.Aiming at the problem of low monitoring accuracy caused by observation noise,a progressive expansion model based on mixed integer programming theory is proposed.The method includes a preliminary solution algorithm and a two-factor variance expansion algorithm.Firstly,its theoretical model is derived.The preliminary solution is updated by the results of three optimal equations The two-factor variance inflation algorithm maintains the symmetry and correlation of the original covariance matrix by adjusting the variance factors of the two equations,and effectively controls the influence of abnormal observations on the parameter estimation.Compared with the traditional algorithm,the model can reduce the noise influence in the observation equation effectively and improve the solution result.(3)According to the principle of partial-space constraint,the prior information in bridge deformation monitoring is transformed into ambiguity search constraints.An MLAMBDA algorithm with partial-space constraints is proposed.The applicability of the theory is proved,and the solving process of the algorithm is deduced.Compared with the traditional algorithm,the proposed algorithm will retain the correct ambiguities while reducing the dimension of the ambiguity candidates.And the proposed algorithm improves the results effectively.(4)The manuscript analyzes the influence of occlusion and noise through dataquality check.The low redundancy of observation equation will result in low-accuracy solutions.Aiming at the problem,a KKT estimation model with the partial ambiguity algorithm constraints is proposed.The method includes partial ambiguity algorithm and supplementary optimal solution algorithm.It is assumed that the partial ambiguity algorithm solution has the same double-difference integer ambiguity as the final solution.The KKT model is to estimate the ambiguities of the other observation equations.Compared with traditional models,the proposed model has more usable observation equations and effectively improves the solution accuracy.(5)The manuscript synthesizes the preceding algorithm into a whole,and analyzes the frequency spectrum of its solutions.On the basis of the vibration characteristics of the bridge,the manuscript applies a nonlinear autoregressive neural network with external input for bridge monitoring time series modeling.This model could predict the time series change continuously,which is essential for bridge health monitoring.