Settlement Prediction Model of Soft Soil Subgrade based on Bayesian Quantile Estimation—Polynomial taking the observe time as the argument was selected as the prediction model of the time data for the time series of settlement of soft soil subgrade with large fluctuation or abnormal points.The polynomial model was easy to be linearized,and the parameters of the model could be estimated by using Bayesian quantile.Example analysis showed that the Bayesian Settlement Prediction Model of Soft Soil Subgrade based on Bayesian Quantile Estimation-Polynomial taking the observe time as the argument was selected as the prediction model of the time data for the time series of settlement of soft soil subgrade with large fluctuation or abnormal points. Polynomial model was easy to be linearized, and the parameters of the model could be estimated by using Bayesian quantile. Example analysis showed that the Bayesian