摘要
根据大坝裂缝开度实测资料的特点,将裂缝开度{xt}看成一系列时刻t1,t2,…,tn得到的时间序列,采用基于滑动平均模型MA(q)的非平稳时间序列法,对其进行模型识别、参数估计和新息预报。通过概率分布函数对其规律性进行统计描述,克服了常规回归分析方法的不足。计算结果表明,该方法预报精度高,是一种行之有效的方法。
On the basis of the characteristics of dam crack data, the crack openings is considered as a time series. The moving average model-based nonstationary time series method is used in the model distinction, parameter estimation, and new information prediction in this paper. This method overcomes the disadvantages existing in the regression analysis with the statistic analyzing. The computation results show that this method is reasonable in the predication of the crack openings, and the prediction reliability is good.
出处
《水电自动化与大坝监测》
2003年第1期47-50,共4页
HYDROPOWER AUTOMATION AND DAM MONITORING
基金
国家自然科学基金重点资助项目(50139030)
河海大学科技创新基金资助项目(2002404443)。
关键词
大坝
裂缝
非平稳时间序列
裂缝开度
滑动平衡模型
dam crack data analysis
nonstationary time series
crack openings
moving average model
new information prediction
