Forecast of coal consumption in Gansu Province based on Grey-Markov chain model

• Published in: Energy (Oxford) 2020-05, Vol.199, p.117444, Article 117444
• Main Authors: Jia, Zong-qian, Zhou, Zhi-fang, Zhang, Hong-jie, Li, Bo, Zhang, You-xian
• Format: Article
• Language: English
• Subjects: Accuracy; Coal; Coal consumption; Coal industry; Consumption; Decision making; Development strategies; Error correction; Forecasting; Fossil fuels; GM(1,1) model; Grey-Markov chain model; Industrial development; Markov analysis; Markov chains; Mathematical models; Model accuracy; Model testing; Model testing, scenario analysis; Predictions
• ISSN: 0360-5442; 1873-6785
• DOI: 10.1016/j.energy.2020.117444

Coal is not only the most important primary energy in China,but also the most stable and safe energy. Accurate prediction of coal consumption can promote the adjustment of the coal industrial structure, accelerate the coal industry to achieve high-quality development,it also provides an effective decision-making basis for the formulation of medium-and long-term coal industry development strategy. Therefore, the prediction of coal consumption has become extremely essential and urgent. In this paper,the coal consumption of Gansu in the past 20 years from 1999 to 2018 is taken as the basic data. First,the GM (1,1) forecasting model of coal consumption in Gansu was established. Based on the forecast of the coal consumption of Gansu in the past two decades,the accuracy of the model was tested. The results show that the predicted average relative error was 0.08881,and the GM(1,1) model was barely qualified with better forecasting accuracy, which is suitable for medium and long-term coal consumption forecast. Subsequently,the Markov chain prediction method was adopted to correct the GM (1,1) model,and the accuracy of the modified model was tested. The results indicate that after correcting the GM (1,1) model with Markov chain,the predicted average relative error was 0.04454,far less than before the correction,and the predicted accuracy was significantly enhanced. Finally,the Grey-Markov chain model was employed to predict the coal consumption in Gansu from 2020 to 2035,and the forecasting process was also analyzed. In addition, choose to use different scenarios to predict coal consumption, then the prediction results obtained by the two methods are compared. •Grey- Markov chain model suitable for coal consumption prediction.•Although renewable resources have great potential, coal still plays a leading role.•Coal has the potential to become a clean and stable source of energy.•Clean and efficient utilization is the development direction of coal industry.•Coal consumption in Gansu will continue to rise slowly in a long period of time.