Gray Time-Varying Parameters Model For Settlement Forecast Of Soft Clay Subgrade

Calculation and prediction models for soft clay settlement have been putting forward constantly since the consolidation theory of Terzaghi was presented. However, their short points are the same: the constant parameters. On the other hand, the fact is that the parameters vary with the development of soft clay consolidation. Theories are so faraway from the fact that lower prediction precision, especially for medium-term and long-term, is brought up. In order to upgrade the prediction precision, the emphasis of this paper is laid one the time-varying property. And a new model- Varying Parameters Model- for settlement prediction is built.Gray Time-Varying Parameters Model is based on the Gray Theory, including equal time-step model and unequal time-step model. In order to upgrade the prediction precision, a forgetting factor is led in, which contributes to the direct modifications of TVP and the indirect prediction results. And its optimum value depends on the posterior square ratio c.Taking the project of XiangJin Expressway as an instance, the settlements of the representative cross-sections are predicted through the Gray TVP Model. According to the results of error checking, the forgetting factor is adjusted to upgrade the error checking level. Based on the analysis of the predicted results of hyperbolic curve model and exponent curve model, the Gray TVP Model gains an advantage over them two in prediction precision and time effect. So Gray TVP Model has a substantial applicable foreground in practical engineering.