基于统计模型与混沌理论的大坝安全监测混合预测模型

为解决传统统计模型在大坝安全监测领域应用中由于缺少对残差的分析而存在拟合精度高但预测效果欠佳的问题，在传统统计模型中增加残差预测项，并应用混沌理论对残差预测项的计算方法进行探讨，从而提出了一种新的混合预测模型，同时结合实测数据对此模型进行了检验。结果表明，新模型可提高预测精度并具有一定的实用性。

Statistic model and chaos theory-based hybrid forecasting model for dam safety monitoring

The conventional statistic model is widely used in the field of dam safety monitoring. Owing to the lack of analysis of residual errors, the forecast results are not satisfactory even though the fitting precision of the model is high. Aiming at the above problem, a hybrid forecasting model is developed based on the chaos theory by introduction of the residual error predicting factor into the conventional model and analysis of its calculating method. The model is tested by field data, and the accuracy and applicability of the model are verified.