Abstract: In order to accurately predict the highway freight trend in China, combining GM (1,1) prediction model, Markov theory and metabolism, a combination forecasting model is proposed. Based on the original data series of highway cargo turnover in China from 2009 to 2016, the highway cargo turnover in China from 2017 to 2019 is predicted. The results show that the prediction accuracy of the combined model is higher than that of the traditional grey GM (1,1) model. By adding the metabolic thought, deleting the old data and introducing the new data, the error of the long-term prediction is reduced. When the grey Markov model is used for the new sequence, the relative error from the original 7.81% and 6.45% is reduced to 3.85% and 0.62%, respectively, during 2018-2019.

Key words: Highway freight traffic; GM (1,1) model; Markov model; Metabolism

摘要： 为准确预测我国高速公路货物运输趋势，文章提出灰色 GM（1,1）模型、马尔科夫模型和 新陈代谢思想的组合模型，以 2009—2016 年我国高速公路货物周转量为原始数据序列，预测 2017 —2019 年高速公路货物周转量。结果表明：组合模型比传统的灰色 GM（1,1）模型预测精度更高， 加入新陈代谢思想，删除旧数据，引入新数据，降低了长期预测的误差，对新序列采用灰色-马尔 科夫模型，2018 年和 2019 年的相对误差由原来的 7.81%和 6.45%分别下降到 3.85%和 0.62%。

关键词: 高速公路货物运输量；灰色 GM（1,1）模型；马尔科夫模型；新陈代谢