基于时间序列的地区能源消耗建模分析与预测

Time-Series-Based Modeling Analysis and Prediction of Regional Energy Consumption

摘要: 能源消耗已成为破坏生态环境平衡的主要因素之一。对城市能源消耗的分析与预测可以为区域差异化能源政策制定及节能减排目标实现提供科学依据。本文以2010~2022年全国及上海市、辽宁省、内蒙古自治区的能源消耗的逐年数据为研究样本。数据经过异常值剔除和插值处理后，采用ARIMA、二次移动平均、Holt线性指数平滑和灰色预测模型对时间序列特征进行分析，并对未来4期能源消耗量进行预测。结果显示，模型的平均相对误差都小于5.2%，可知模型预测精度较好。此外，研究表明，上海市、辽宁省和内蒙古自治区的煤炭等传统高碳能源消耗在总消费中的占比呈逐步下降趋势。

Abstract: Energy consumption has become one of the main factors disrupting the balance of the ecological environment. The analysis and prediction of urban energy consumption can provide a scientific basis for the formulation of regionally differentiated energy policies and the achievement of energy conservation and emission reduction goals. This paper takes the annual energy consumption data of the whole country, Shanghai Municipality, Liaoning Province, and the Inner Mongolia Autonomous Region from 2010 to 2022 as research samples. After outliers removal and interpolation processing of the data, ARIMA, quadratic moving average, Holt linear exponential smoothing, and grey prediction models are used to analyze the characteristics of the time series and predict the energy consumption for the next 4 periods. The results show that the average relative error of all models is less than 5.2%, indicating good prediction accuracy. In addition, the study shows that the proportion of traditional high-carbon energy consumption such as coal in the total consumption of Shanghai Municipality, Liaoning Province, and the Inner Mongolia Autonomous Region is showing a gradual downward trend.