关键词: 沉降预测, 时变参数, 遗忘因子, 误差检验, 灰色模型

Abstract: The settlement prediction of soft ground is a typical time varying system, of which the parameters have time-varying property; hence, the time-varying property of parameters taken into account in the grey model for settlement prediction is significant. According to the information handling principle of the grey theory, the unequal step lengths grey time-varying parameters model for settlement prediction has been built. In the model, both time-varying and grey property　are considered; and a forgetting factor, of which the optimum value is validated according to the posterior variance ratio c, is quoted in the model to correct prediction results. The example shows that the unequal step lengths grey time-varying parameters model for settlement prediction has applicability, and prediction error can be decreased and the prediction precision can be enhance by the grey time-varging paraheters model.

Key words: settlement prediction, time-varying parameters, forgetting factor, error checking, grey model