摘要
针对大坝实时性态预测方法精度,首先基于Ito随机微分方程对某坝的多年扬压力监测极值建立GM(1,1)了实测序列预测模型,并根据建立模型的残差序列构建了Markov残差模型,其次对比分析了GM(1,1)残差预测模型和Markov残差模型。最后,综合GM(1,1)模型和Markov残差模型建立灰色Markov模型,并利用灰色Markov模型预测扬压力的极大值。计算结果表明,建立的灰色Markov模型不仅能提高预报精度还能真实地反映其过程的摆动性。
Aiming at the problem of accuracy of real time state prediction method. First of all, based on the Ito stochastic differential equation, the GM (1, 1 ) prediction model is established for monitoring the extreme pressure of a dam, and the Markov residual model is constructed according to the residual sequence. Secondly, the GM (1,1) residual prediction model and Markov residual model are compared and analyzed. Finally, the GM (1,1) model and the Markov residual model are used to establish the grey Markov model, and the maximum value of the uplift pressure is predicted by using the grey Markov model. The results show that the grey Markov model can not only improve the prediction accuracy, but also truly reflect the swing of the process.
作者
郭岩虎
GUO Yan-hu(Changji Water Conservancy Bureau Key Water Conservancy Project Construction Administrative Office, Changji 831100, China)
出处
《水科学与工程技术》
2017年第3期91-94,共4页
Water Sciences and Engineering Technology
关键词
Ito随机微分方程
灰色Markov模型
残差辨识
关联度
预报精度
Ito stochastic differential equation
grey Markov model
residual identification
correlation degree
prediction accuracy