摘要
针对各预报模型预报结果精度评价不统一的现状,考虑径流具有非线性、突变及非平稳性等特点,本文构建了包含均方根误差(RMSE)、平均绝对百分误差(MAPE)和Nash效率系数(NSE)三项指标的综合评价系统,对自回归滑动平均模型(ARMA)、人工神经网络模型(ANN)和支持向量机模型(SVM)在径流汛期和非汛期内进行了预报精度评价。结果表明:①单一评价指标下,ARMA模型与SVM模型预报结果精度相近,而综合评价系统表明,SVM模型预报精度优于ARMA模型;②三种模型在非汛期预报精度均高于汛期预报精度,SVM预报效果均最好。将径流进行分割后预报,预报精度可提高。本研究获得了可靠性和精度较高的月径流预报模型,可为工程水资源高效配置提供理论和技术支撑。
In view of the current situation that the prediction results accuracy evaluation of various prediction models is not uniform,with the characteristics of non-linearity,abrupt change and non-stationary of runoff considered,a comprehensive evaluation system with three indexes is constructed in this paper,which includes the root mean square error(RMSE),the mean absolute percentage error(MAPE)and the Nash efficiency coefficient(NSE).The prediction accuracy of the autoregressive moving average model(ARMA),artificial neural network model(ANN)and support vector machine model(SVM)are evaluated in flood season and non-flood season.The results show that:①Under a single evaluation index,the prediction accuracy of the ARMA model and the SVM model are similar,and the comprehensive evaluation system shows that the prediction accuracy of the SVM model is better than that of the ARMA model;②The prediction accuracy of the three models in non-flood season is higher than that in flood season,and the SVM prediction effect is the best.The forecast accuracy can be improved by dividing the runoff.This study has obtained a reliable and accurate monthly runoff forecast model,providing the theoretical and technical support for the efficient allocation of engineering water resources.
作者
李静
黄强
杨元园
黄生志
刘登峰
孟二浩
LI Jing;HUANG Qiang;YANG Yuanyuan;HUANG Shengzhi;LIU Dengfeng;MENG Erhao(State Key Laboratory of Eco-hydraulics in Northwest Arid Region of China,Xi’an University of Technology,Xi’an 710048,China;Shaanxi Key Laboratory of Water Resources and Environment,Xi’an 710048,China)
出处
《西安理工大学学报》
CAS
北大核心
2021年第3期338-344,共7页
Journal of Xi'an University of Technology
基金
国家自然科学青年基金资助项目(52009099)
国家自然科学面上基金资助项目(51779203)
中国博士后科学基金资助项目(2019M653882XB)
陕西省教育厅基金资助项目(18JS074)
清华大学水沙科学与水利水电工程国家重点实验室联合基金资助项目(sklhse-2019-Iow06)。
关键词
径流预报
支持向量机模型
综合评价系统
引汉济渭工程
汛期与非汛期
runoff forecast
support vector machine model
comprehensive evaluation system
the Han to the Wei diversion project
flood season and non-flood period
