基于灰色 Markov模型的我国远洋船舶货运总量预测*

摘要

The forecast of the total freight of Chinese ocean shipments relates not only to a series of scientific project of national port construction and shipping development ,but also to the proportion of ocean shipping in integrated transport as well as the resource allocation of international maritime transport .The simple use of grey model always predicts well for ob‐vious monotonous sequence ,but for volatility sequences ,the problem of unsatisfactory accuracy occurs .Combining the ad‐vanced grey prediction model and the theoretical advantage of Markov model ,this article firstly shows the prediction of over‐all trend of the Chinese ocean shipments with the improved grey model ,then depicts the way of transfer between the various states ,so as to improve the prediction accuracy of volatility sequences ,and is of great practical value .%对我国远洋船舶货运总量的预测关系到我国港口建设和航运发展的一系列科学规划，也关系到远洋运输在综合交通运输中的比重和国际海上运输的资源分配。单纯的灰色模型对明显单调的序列往往预测得较好，而对波动的序列则会出现精度不够理想的问题。在结合改进后的灰色预测模型和M arkov模型的理论优势后，先用改进后的灰色模型来预测远洋船舶货运总量的总体走势，再用M arkov模型来确定各种状态间的转移方式，以此提高对波动序列的预测精度，具有较高的实用价值。