摘要
海上交通事故预测具有原始数据量小、数据随机波动大的特点。在传统的灰色预测模型GM(1,1)基础上,运用系统云灰色模型SCGM(1,1)c预测事故数,并用马尔可夫链对预测值进行修正。同时,为了使预测结果更加接近真实值,采用等维灰数递补法和滑动转移概率矩阵法对组合模型进行改进。结果表明,双重改进后的组合模型能够克服数据随机波动的影响,并且比单一改进的组合模型的预测结果更精确。
The prediction of maritime traffic accidents faces the problems of small original data set and great data fluctuation.On the basis of traditional grey forecasting model,this paper forecasts accidents quantitatively by using SCGM（1,1）c corrected according to Markov chain.The combined model is refined by using the methods of the equal dimensional gray recurrence and progressive transition probability matrix.Results show that the double improved combination model can overcome the effect of the random fluctuation of data and give predictions more precise than SCGM（1,1）c with correction.
出处
《中国航海》
CSCD
北大核心
2013年第4期119-124,共6页
Navigation of China
关键词
水路运输
海上交通事故
系统云灰色模型
马尔可夫链
改进的
预测
waterway transportation
maritime traffic accidents
system cloud grey model
Markov chain
improved
forecast
