基于时间序列分析的云南省能源消费预测研究

Research on Energy Consumption Prediction of Yunnan Province Based on Time Series Analysis

摘要: 本文应用时间序列分析方法的理论基础，结合云南省1976年至2020年能源消费总量数据为样本，利用R软件对能源消费总量建立了ARIMA模型和Holt两参数指数平滑模型，将预测到的2016年至2020年能源消费总量的预测数据与其实际值做对比，根据相对误差绝对值取最小的选择原则，发现Holt两参数指数平滑模型预测的数据更加贴近实际值，于是将其作为最优模型进行预测。根据预测结果得出结论：2021年至2025年云南省的能源消费总量呈直线上升趋势。

Abstract: Based on the theoretical basis of the time series analysis method and the total energy consumption data of Yunnan Province from 1976 to 2020 as samples, this paper uses R software to establish ARIMA model and Holt two-parameter exponential smoothing model for the total energy consumption, compares the predicted total energy consumption data from 2016 to 2020 with the actual value, and takes the minimum absolute value of relative error as the selection principle, It is found that the data predicted by Holt two-parameter exponential smoothing model is closer to the actual value, so it is used as the optimal model for prediction. According to the prediction results, it is concluded that the total energy consumption in Yunnan Province will rise linearly from 2021 to 2025.