摘要
针对秩次集对分析方法在年径流预报中存在的问题,对其历史集合中的各元素所在位置与后续值的位置之间的间隔赋予了不同的影响权重,构建了年径流加权秩次集对分析预报模型,给出了模型在年径流预报应用中的步骤,利用黑河干流莺落峡站1945～2009年的实测径流资料检验了该模型的有效性,并与秩次集对预报模型、ARIMA预测模型及BP预测模型的预测结果进行了比较。结果表明,加权秩次集对预报模型的平均误差为12.29%、合格率为100%,本文对秩次集对预报模型的改进有效、可行。
In order to improve the rank set pair analysis method for forecasting annual runoff, different impact weight is given to the different time interval of the locations between each element in the historical set and the subsequent ele- ment. And the weighted rank set pair analysis model is developed and the procedure for application of the model to annual runoff forecasting is presented. Based on the annual runoff data of Yingluoxia hydrological station in Heihe river from 1945 to 2009, the weighted rank set pair analysis forecasting model is verified, and compared with rank set pair analysis forecasting model, ARIMA forecasting model and BP forecast model. The results show that the average relative errors and the qualified rate of the proposed forecasting model are 12.29%and 100%, respectively： improvement of the weighted rank set pair analysis forecasting model is effective and feasible.
出处
《水电能源科学》
北大核心
2012年第3期17-19,67,共4页
Water Resources and Power
基金
国家自然科学基金资助项目(40830639)
关键词
年径流
水文预报
集对分析
加权
预报模型
annual runoff
hydrological forecasting
set pair analysis
weight
prediction model
