贴近度的诱导有序加权平均算子最优组合模型

Application of combined model based on approach degree and induced ordered weighted averaging operator on subgrade settlement

针对由路基沉降观测数据建立的各种沉降预测模型各有优缺点的问题,该文提出了将基于贴近度的诱导有序加权平均算子最优组合模型应用于路基沉降变形预测分析中的方法。该文将贴近度与诱导有序加权平均算子进行结合,建立基于贴近度的诱导有序加权平均算子最优组合预测模型,并用于沉降预测分析。对各模型的预测精度进行比较分析,结果显示:该组合模型效果优于单项模型的预测精度,预测相对误差小,最大误差<5%;其各项误差指标也低于参与组合的各单项预测方法。

Because all the prediction models of the subgrade settlement based on observation data have both advantages and disadvantages, combined model based on approach degree and induced ordered weighted averaging（IOWA）operator on subgrade settlement was given in this paper. The combination forecasting model was established, which was based on approach degree and IOWA operator optimal for settlement prediction and analysis. The predicted accuracy of each model was compared. The results showed that the predicted accuracy of the new combined model performed better than the individual models＇. The forecasting relative error was small, and the maximum error was less than 5%, its error indicators were also lower than the individual forecast method.