加权多变量灰色模型在变形监测中的应用

在变形监测过程中,监测因子较多,难以对监测因子进行合理的取舍。首先利用灰关联度分析方法,将监测因子根据灰关联度进行关联排序,为合理选择监测因子提供理论依据;然后,利用得到的关联度进行平滑处理,并将其作为权值建立加权多变量灰色模型,推广了传统的多变量灰色模型,提高了预测精度;最后,以南充水库土石坝沉降数据为例,验证了模型的正确性。算例结果表明,加权多变量灰色模型的平均相对误差比传统灰色模型小,模型精度也比传统模型高。

Application of Weighted Multi-variable Grey Model in Deformation Monitoring

In the process of deformation monitoring,there are more monitoring factors.It is very difficult to reasonably choose from the monitoring factors.First,in this paper,we used the grey correlation analysis method,sorted the monitoring factors based on the grey relational grade for provide a theoretical basis for a reasonable choice of monitoring factors.Second,we smoothed the relational grade what had been gotten,and regarded it as weight to establish weighted grey model.Not only spread the traditional multi-variable grey model,but also improve the prediction accuracy.In the last of this paper,we verified the model based on the Nanchong reservoir embankment dam settlement data.The result shows that the average relative error of the weighted multi-variable grey model is smaller than traditional model,and the precision of the weighted multi-variable grey model is higher than traditional model.