基于灰色马尔可夫链模型的中国能源消费预测研究  

Forecast Research into Chinese Energy Exprense Based on Gray Markov China Model

作　　者：刘亮亮[1] 敖军[1] 高世泽[1] 

机构地区：[1]重庆师范大学数学与计算机科学学院,重庆400047

出　　处：《重庆师范大学学报（自然科学版）》2008年第4期47-49,69,共4页Journal of Chongqing Normal University:Natural Science

摘　　要：随着中国经济的不断发展,能源需求不断上升,能否准确预测中国能源消费量对中国经济乃至全球经济发展都有重要的指导意义。现行的预测方法主要有回归分析法、经验模型法、时间序列法、指数平滑法和灰色系统预测法等。灰色系统预测方法在预测波动性较大的非平稳数列上有不足之处,而马尔可夫预测却适用于随机波动性较大的数列的预测。因此本文将灰色系统理论与马尔可夫链相结合,建立了能源消费总量数据的灰色马尔可夫链模型。实证分析表明,这种模型的预测精度高于GM(1,1)预测模型,预测效果较好。With the development of the economy, the demand of energy rises unceasingly. The accurate forecast Chinese sources of energy consumption has important guiding significance to Chinese economy even to the whole world economy development. The gray system needs the less information. The calculation is simple and convenient and the accuracy is higher. So it has been widely applied recently. Grey forecast methods in predicting volatility of the larger non-stationary series are inadequate. Markov is used to forecast random fluctuations of the larger series of forecast. This paper is based on grey system theory and Markov chain integration. The establishment of the total energy consumption data with grey Markov chain model. At first, the paper utilizes the gray system to get the prediction of every year. Secondly, it compares the value of effect with the value of prediction to get the ratio and then to get receive the states. By the state of the changes we get the transfer matrix. Lastly, we use the Markov chain to predict. Empirical analysis shows that this model of prediction has accuracy higher than that of GM （ 1, 1 ） prediction model. This method is also suitable for other forecast.

关 键 词：能源消费 灰色系统 GM(1 1)模型 马尔可夫链 

分 类 号：O21[理学—概率论与数理统计]