在建筑物施工及运营期间,掌握建筑物变形规律并及时预测其变形趋势在保障建筑物的安全使用方面发挥着重要作用。由于单一的灰色GM(1,1)模型预测精度不高,本文提出了一种预测模型——灰色自回归组合模型,该模型结合了灰色GM(1,1)模型和自回归模型的优点。为验证该模型预测精度优于灰色GM(1,1)模型,本文通过某工程实例中的2个变形观测点的观测数据进行建模,结果表明,组合模型的预测精度高于单一灰色GM(1,1)模型。 During the construction and operation of a building, it is important to master the deformation rules of buildings and predict the deformation trend in time in order to ensure the safe use of buildings. Since the prediction accuracy of a single gray GM (1,1) model is not high, this paper proposes a combined model of gray autoregressive model, which combines the advantages of gray GM (1,1) model and autoregressive model. In order to verify that the prediction accuracy of this model is better than that of gray GM (1,1) model, this paper builds a model by observing data of 2 deformation observation points in an engineering example. The results show that the prediction accuracy of the combined model is higher than that of single gray GM 1,1) model.