基于小波-ARIMA的航空运输企业能耗预测模型

摘要

In order to forecast the monthly energy consumption of aviation transportation enterprise,the hybrid forecasting model based on wavelet analysis and ARIMA is proposed.Based on the characteristics of the monthly energy consumption sequence of the aviation transport enterprise,the sequence of energy consumption is decomposed into two dimensions,and then which are forecasted by the model of ARIMA,finally the forecasting results are reconstructed.The results with the data from January 2009 to December 2013 of an aviation transportation enterprise show that the forecasting accuracy of the wavelet-ARIMA model is higher than the model of ARIMA,and the model of the improved wavelet-ARIMA improve the comprehensive forecasting accuracy.%针对航空运输企业月度能耗进行预测,提出基于小波分析与求和自回归移动平均模型(ARIMA)相结合的混合预测模型.该模型根据航空运输企业月度能耗序列的特征,对其进行二尺度小波分解,对分解后的能耗序列根据自身规律分别进行ARIMA预测,重构各尺度序列的预测值得到最终预测结果.利用莱航空运输企业2009年1月至2013年12月五年的月度能耗数据进行验证,结果表明:小波-ARIMA模型的预测精度高于ARIMA模型,对数据的波动性有较好的响应,提高了能耗序列的整体预测精度.