摘要
预测模型的准确与否直接决定着未来经济规划与决策的有效制定。将灰色GM(1,1)-Verhulst组合预测模型与马尔可夫链方法相结合,同时引入信息熵理论的知识,提出基于Markov链修正的熵权法灰色组合预测方法,并以甘肃省2004年～2015年铁路客运量作为原始数据序列进行模型拟合,而且还以此为基础对甘肃省未来几年内的客运量发展趋势进行预测。结论:(1)在已知实际客运量年份内,该灰色组合预测模型的预测精度比单一灰色预测模型更高、更加准确;(2)采用马尔可夫链方法获得该组合模型的偏差规律,并依照此规律对预测结果进行修正,即由一个单一的预测数值修正成为区间和概率组成的预测范围;(3)通过比较2016年～2017年的客运量实际值、组合预测模型的单一预测值和Markov链修正的预测区间值,发现Markov链修正的预测结果与客运量实际值的吻合性良好,进一步验证此预测方法的可信性。
The accuracy of the prediction model directly determines the effective formulation of future economic planning and decision-making.Therefore,this paper combines the grey GM(1,1)-Verhulst combination forecasting model with the Markov chain method,introduces the knowledge of information entropy theory,and proposes a gray combination forecasting method based on the Markov chain correction and entropy weighting method.The railway passenger traffic in Gansu Province from 2004 to2015 is modeled as the original data series,and on the basis of which the development trend of passenger traffic in Gansu Province in the nextfew years is forecasted.It is concluded that:(1)in the actual year of known passenger traffic,the prediction accuracy of this grey combination forecast model is higher than that of a single gray forecast model;(2)the Markov chain method is used to obtain the deviation rule of the combined model,and the prediction results are modified according to this rule,that is,a single prediction value is modified to form the prediction range of the interval and probability;(3)by comparing the actual values of passenger traffic,the single predictive value of the combined forecasting model,and the forecasted interval value modified by the Markov chain from 2016 to 2017,it is found that the Markov chain’s revised forecasting result is in good agreement with the actual value of passenger traffic,and the credibility of this prediction method is further proved.
作者
贾鼎元
柴乃杰
王恩茂
JIA Ding-yuan;CHAI Nai-jie;WANG En-mao(Institute of Civil Engineering,Lanzhou Jiaotong University,Lanzhou,730070,China)
出处
《铁道标准设计》
北大核心
2019年第2期25-30,共6页
Railway Standard Design
基金
国家自然科学基金(51768034)
长江学者和创新团队发展计划滚动支持(IRT1139)
关键词
铁路客运量
经济预测
灰色预测
MARKOV链
熵权法
组合预测
Railway freight volume
Economic forecasting
Grey prediction
Markov chain
Entropy weight method
Combined forecasting
