信息扩散方法在大坝监测效应量预测中的应用

在大坝安全监测中,为了解决小样本导致的效应量预测误差较大的问题,在考虑增加样本信息的基础上,建立基于二维正态信息扩散离散回归的效应量预测方法。利用模糊推理和信息集中的基本理论,结合灰色关联度法,对大坝安全监测效应量进行分析和预测,应用综合有效度指标评价该预测方法的效果。结果表明,基于信息扩散的预测方法避免了多种不确定因素影响下小样本预测的困难,提高了拟合和预测精度。

Application of information diffusion approach to effect quantity prediction in dam monitoring

In order to solve the problem of large errors in effect quantity prediction caused by small samples in dam safety monitoring, a new approach for prediction of effect quantity, through increasing sample information, was established based on discrete and regression analysis of two-dimensional normal information diffusion. The effect quantity of dam safety monitoring was analyzed and predicted using the basic theory of fuzzy reasoning and information concentration, as well as the grey correlation degree method. The comprehensive effectiveness index was used to evaluate the effect of the information diffusion approach. The results show that the information diffusion approach avoids the difficulties in effect quantity prediction using small samples under the influence of some uncertain factors, and improves the fitting and forecast accuracies.