面向统一能源系统的中长期氢负荷预测

Medium and Long-Term Hydrogen Load Forecast for Unified Energy System

未来能源社会中氢气将在电力、工业、供热、交通等领域发挥巨大作用,氢能将作为统一能源系统的关键要素,实现各能源相互转化。针对未来社会中氢能在工业、供热、交通等领域的需求,提出一种氢负荷预测的方法。获取工业领域的氢负荷样本数据,算出负荷数据的特征,采用支持向量机回归(SVR)算法,得到工业领域氢负荷预测模型;然后,以供热、交通领域需氢数据建立模型,采用改进灰色GM(1,1)模型与新陈代谢模型结合,得到供热、交通领域氢负荷预测模型;最后,叠加3种氢负荷预测,完成数学模型构建。从结果可以看出SVRT预测方法十分准确、改进灰色GM(1,1)模型与新陈代谢的组合模型组预测精度较高,该方法可用于中长期氢负荷预测。

In the future energy society, hydrogen will play a huge role in power system, industrial application, heating supply,transportation and other fields. Hydrogen energy will be used as the key element of the unified energy system to achieve the energy conversion between various energy sources. In this paper, a method of hydrogen load forecast is proposed for the future demand of hydrogen energy. Firstly, on the basis of the hydrogen load sample data from the industrial field, the characteristics of the load data are calculated and the support vector machine regression(SVR) algorithm is applied to set up the hydrogen load forecast model accordingly. Secondly, a model based on the hydrogen demand data in the heating supply and transportation fields are built as well,and the improved gray GM(1,1) model is utilized in combination with the metabolic model to obtain a hydrogen load forecast model in the heating supply and transportation fields. Finally, all above three hydrogen load forecast are superimposed together to complete the construction of the mathematical model. From the results, not only the SVRT prediction method is proved to be very accurate, but also the combined model group based on the improved gray GM(1,1) model and metabolism have demonstrated high forecast precision. This method is feasible for medium and long-term hydrogen load prediction.