摘要
简要介绍了组合预测模型的基本原理,该模型改进了单一模型考虑因素的片面性,但常规的组合预测模型的权系数是固定的。考虑到各种模型的预测精度的变化,以及大坝安全监测数据存在的突变、时效等因素,固定权系数的组合预测模型有待进一步改进。文中提出了变权系数组合预测模型,并将其用于实际的大坝安全监测工程。结果表明,变权组合预测模型的预报精度高于其他模型,是一种值得推广的方法。
The principle of a combination forecasting model is briefly introduced. This model corrects the one-sidedness of single models in considering influence factors, while the weights of a conventional combination forecasting model are fixed. Considering the precision of different models, the abrupt change and time effect of dam safety monitoring data, a combination forecasting model with variable weights is proposed and applied to a practical dam safety monitoring project. The result shows the combination forecasting model with variable weights is more accurate and efficient.
出处
《水电自动化与大坝监测》
2006年第5期60-62,共3页
HYDROPOWER AUTOMATION AND DAM MONITORING
基金
国家自然科学基金(50579010)
国家重点基础研究发展计划(973计划)资助项目(2002CB412707)
关键词
大坝安全监测
组合预测模型
变权系数
dam safety monitoring
combination forecasting model
variable weight