基于EEMD-AR模型的丹江口水库年径流随机模拟与预报
发布时间：2018-11-03 13:10
【摘要】：基于水库历史年入库径流序列组分分析和识别,采用线性趋势回归检验法、有序聚类法、方差线谱法等方法,推求出序列趋势项、跳跃项及周期项等确定性成分,提出基于集合经验模态分解法(EEMD方法)的水库年径流自回归随机模拟模型(EEMD-AR),并应用于丹江口水库的年径流随机模拟和预报中。通过EEMD分解,解决了当丹江口水库历史年径流序列为非平稳序列时不能直接应用自回归模型(AR)进行随机模拟和预报的问题。模拟结果表明,EEMD-AR模型能较好地模拟丹江口水库年径流序列并保持原历史序列的统计特性,且模型预报精度符合要求。
[Abstract]:Based on the analysis and identification of the components of the runoff sequence in the historical year of the reservoir, the deterministic components such as the sequence trend term, the jump term and the period term are derived by using the linear trend regression test method, the ordered clustering method and the variance line spectrum method. An annual runoff autoregressive stochastic simulation model (EEMD-AR) based on the set empirical mode decomposition (EEMD) method is proposed and applied to the stochastic simulation and prediction of the annual runoff of Danjiangkou reservoir. Through EEMD decomposition, the problem that the autoregressive model (AR) can not be directly used for stochastic simulation and prediction when the historical annual runoff series of Danjiangkou reservoir is non-stationary is solved. The simulation results show that the EEMD-AR model can simulate the annual runoff series of Danjiangkou reservoir and maintain the statistical characteristics of the original historical series, and the prediction accuracy of the model meets the requirements.
【作者单位】： 天津大学水利工程仿真与安全国家重点实验室;
【基金】：国家重点研发计划水资源高效利用专项(2016YFC0402203)
【分类号】：P333;P338
[Abstract]:Based on the analysis and identification of the components of the runoff sequence in the historical year of the reservoir, the deterministic components such as the sequence trend term, the jump term and the period term are derived by using the linear trend regression test method, the ordered clustering method and the variance line spectrum method. An annual runoff autoregressive stochastic simulation model (EEMD-AR) based on the set empirical mode decomposition (EEMD) method is proposed and applied to the stochastic simulation and prediction of the annual runoff of Danjiangkou reservoir. Through EEMD decomposition, the problem that the autoregressive model (AR) can not be directly used for stochastic simulation and prediction when the historical annual runoff series of Danjiangkou reservoir is non-stationary is solved. The simulation results show that the EEMD-AR model can simulate the annual runoff series of Danjiangkou reservoir and maintain the statistical characteristics of the original historical series, and the prediction accuracy of the model meets the requirements.
【作者单位】： 天津大学水利工程仿真与安全国家重点实验室;
【基金】：国家重点研发计划水资源高效利用专项(2016YFC0402203)
【分类号】：P333;P338
【相似文献】
相关期刊论文 前10条
1 刘福平,陈玉祥,杨长春;河道油藏边界面的最优化随机模拟[J];地球物理学进展;2002年04期
2 胡康萍;;枯水径流过程的随机模拟研究[J];水电能源科学;1985年04期
3 王文圣,丁晶,邓育仁;日流量过程随机模拟模型的探讨[J];四川联合大学学报(工程科学版);1997年04期
4 贺崇铃;王燕生;;降雨过程随机模拟[J];内蒙古水利科技;1987年03期
5 方乐润;;红水河径流的随机模拟研究综述[J];河海科技进展;1991年04期
6 G萨克里斯;黄炎初;;连续时间内降雨发生的随机模拟[J];广西水利水电科技;1990年03期
7 王文圣,丁晶,袁鹏;非参数解集模型及其在水文随机模拟中的应用[J];四川水力发电;1999年01期
8 邓育仁，，丁晶;间断雨量序列随机模拟的研究──暴雨洪水流域系统随机模拟研究之四[J];成都科技大学学报;1994年05期
9 丁晶;;正则展开模型在径流随机模拟中的应用[J];成都科技大学学报;1988年02期
10 王达雨,张耀宾,李望安,侯云青,林小丽;龙滩水电站入库洪水随机模拟研究[J];水力发电;2004年06期
相关会议论文 前1条
1 鲍振鑫;刘九夫;翟劭q
本文编号：2307882
本文链接：https://www.wllwen.com/kejilunwen/shuiwenshuili/2307882.html
