鉴于高速公路施工的被动性,利用时间序列建立数学模型。对具有非平稳性的表面沉降进行一次差分(d=1)处理,使其转化为具有平稳性的沉降速率问题,即将非平稳性时间序列ARI MA(p, d, q)模型转化为平稳的ARMA(p, q)模型。建立了ARMA(1 ,1)模型来对公路施工填筑期间的沉降进行短期预测,进而对下一级加载的时间进行预测。结果表明,预测结果可靠,精度高,能够很好地反映填筑期间沉降的变形特性,适用于高速公路重点断面分析。 In view of the passivity of expressway construction, the mathematical model is established by time series. The differential settlement (d = 1) of non-stationary surface subsidence is transformed into a stable sedimentation rate problem, that is, the non-stationary time series ARI MA (p, d, q) ARMA (p, q) model. The ARMA (1, 1) model is established to predict short-term subsidence during road construction and predict the loading time of the next stage. The results show that the prediction results are reliable and high precision, which can well reflect the deformation characteristics of settlement during filling, and are suitable for the key section analysis of freeway.