Time Series InSAR Deformation Estimation For Soft Clay Highway Based On Maxwell-NVPB Rheological Model

One of the important reasons for the uneven settlement,cracking and even collapse of the highway is the post-construction deformation of the expressway,especially for the highway built on the soft soil subgrade.At present,the long-term stability control and disaster prediction of soft soil subgrade highway after construction has become a major difficulty in practical highway engineering.Therefore,it is a great significance to carry out long-term continuous deformation monitoring research for soft soil roadbed roads.Although the traditional deformation monitoring method has the characteristics of high monitoring accuracy,its application in the field of traffic surface engineering monitoring is limited due to the defects of high cost,time-consuming and labor-intensive,and mostly based on discrete point monitoring.Small Baseline Subset InSAR(SBAS-InSAR)is an advanced InSAR technology developed in recent years,which has been proven in the post-construction deformation monitoring of large-scale transportation infrastructure(railway,highway and bridge,etc.).Great potential.However,although the traditional time series InSAR technology can realize the deformation monitoring work in this engineering field,the commonly used pure mathematical empirical model cannot describe the evolution characteristics of the strong nonlinear deformation sequence of the soft soil subgrade accurately and reliably,and its deformation monitoring accuracy is affected to a certain extent..On this basis,combined with rheological theory,this thesis introduces the nonlinear viscoelastic-plastic four-element combined rheological model,Maxwell-NVPB,which takes into account the rheological properties of soft soil,into the traditional time series InSAR technology,and constructs rheological parameters and soft soil.The line-of-sight surface settlement relationship function model of highway subgrade overcomes the limitations of the traditional soft soil linear deformation model that does not consider the deformation mechanism of the monitoring object and has poor modeling accuracy.The main research contents and research results of this thesis are as follows:(1)Combined with the internal mechanism of post-construction surface settlement of soft soil subgrade,a nonlinear Maxwell-NVPB rheological model was introduced to construct rheological parameters and the time series function model between the InSAR differential interference phase;(2)The nonlinear parameter estimation research for the improved combined model is carried out,the GA-SM serial optimization algorithm is introduced,and the simulation experiment is designed to verify the feasibility and reliability of the algorithm.The experimental results show that even with a high noise level of 0.5 rad,the RMSE of the estimated value of each unknown parameter is still less than 4%,which confirms the feasibility and reliability of the parameter estimation algorithm;(3)Based on the improved combination model,the real experiment of some sections of Lungui Road in Guangdong and the application case of the new method on some sections of Yanhu Avenue in Yueyang,Hunan were carried out,and the parameter estimation and time series deformation results of the model were obtained.The results show that the monitoring results of the two survey areas show that the magnitude of each rheological parameter is inversely correlated with the subsidence of the corresponding area,and the overall subsidence rate presents a nonlinear change trend from fast to slow;the subsidence of the Dongting Lake area along the lake It is more obvious,and the maximum cumulative deformation is 46 mm;the relative settlement of inland urban areas is relatively small,and the maximum settlement is 12 mm;(4)By comparing the residual phase index with the traditional linear model,it can be seen that the RMS of the traditional linear model is 0.42 rad,and the RMS of the Maxwell-NVPB model is 0.24 rad,which is 41.9%higher than that of the method in this thesis.Sequence comparison shows that the RMSE of the traditional linear model is 17.2 mm,and the RMSE of the Maxwell-NVPB model is±5.6 mm,and the deformation accuracy obtained by the method in this thesis is improved by 24%.(5)The sensitivity analysis of the unknown parameters of the model was carried out,and the influence of the estimation errors of the parameters in the model on the deformation results was revealed.The rheological parameters are all more sensitive than linear velocity v.Among them,viscosity is the highest sensitivity.It is consistent with the conclusions of this study.