基于灰色系统神经网络的中国低碳能源供需形势分析

Analysis of Low Carbon Energy Supply and Demand Situation in China Based on Grey SystemNeural Network

利用灰关联理论分析低碳能源的供需关系,通过构建GM(1,1)模型对中国低碳能源的供需形势进行分析和预测,并运用BP神经网络对低碳能源的需求量进行进一步的对比预测。研究表明,水电的生产量与消费量的关系最为密切,其次是核电,可再生能源的生产量与消费量的关系相对较弱;核电和水电的供需影响因子构建的GM(1,1)模型与可再生能源相比较为稳定,BP神经网络的预测值要略低于灰色系统的预测值;核电、水电、可再生能源的生产量和消费量都将呈现增长的趋势,核电的消费量要略高于其生产量,这种局面将在2025年得到扭转,水电的生产量和消费量将在未来一段时间稳定、持续增长,可再生能源的消费量将高于其生产量,并且随着时间的推进,这种差距会扩大。

In this paper, we used the theory of grey relation to analyze low carbon energy supply and demand in China, and constructed GM（1, 1） model to analyze and predict China＇s low carbon energy supply and demand situation, and then comparatively predict the low carbon energy demand by using BP neural network. Studies showed that the most closely relationship between production and consumption was water electricity, followed by nuclear power. The relations between renewable energy production and its consumption was relatively weak. Further research found that GM（I, 1） model which was built by supply and demand effect factor of nuclear power and hydroelectric power was more stable, compared with renewable energy. Predictive value of BP neural network was slightly lower than grey system; Nuclear power, hydropower, renewable energy production and consumption will show growth trend, nuclear power consumption is slightly higher than that of its production. This situation will be reversed in 2025, hydroelectric power production and consumption will be stable and sustainable growth in tile future for a period of time. Renewable energy consumption will be higher than its production, and with the advance of time, this gap will expand.