基于灰色马尔科夫模型的煤炭产量预测

Study on Coal Production Prediction Based on Gray Markov Model

煤炭资源是我国经济建设中的重要能源,准确预测我国煤炭产量有助于国家能源政策的制定.为此,在单一的灰色GM(1,1)预测模型的基础上,与马尔科夫预测模型组合形成灰色GM(1,1)-马尔科夫组合预测模型,以相对误差、均方差比值和小概率误差3个指标对模型的精度做检验.结果表明,组合预测模型优于单一预测模型,组合预测模型不仅反映了煤炭产量短中期呈上升趋势,又突显了该模型能较优地处理波动性序列的优点,能够有效地预测2019年和2020年煤炭产量.

Coal resource is an important energy source in China’s economic construction. Accurate prediction of China’s coal output is beneficial to the formulation of national energy policy. Based on the single grey GM(1,1) prediction model, it is combined with Markov Prediction Model to form Grey GM(1,1)- Markov combination prediction model. The relative error mean square ratio and small probability error are used to test the accuracy of the three index models. Results show that the combined forecasting model is superior to the single forecasting model. The combination prediction model not only reflects the rising trend of coal production in the short and medium term, but also highlights the advantage of the model in dealing with the fluctuation series. The predicted coal production for the period 2019~2020 is 38.97 and 39.16 billion tons.