基于神经网络的电力系统节点碳排放因子预测方法

Study on Prediction Method of Node Carbon Emission Factors in Power System Based on Neural Network

电力减碳是实现国家“双碳”目标的关键环节。基于碳流分析法的电碳计量方法,能较好地反映发电、电网和用户的真实碳排放情况。然而,碳流分析法的计算量大,时间滞后,难以支撑电力低碳调度和低碳生产等。针对此,提出基于Dropout神经网络的电力系统节点碳排放因子预测方法,并在3个IEEE标准节点系统上,将该方法与传统神经网络、潮流计算进行性能对比。结果表明:该方法准确度高,计算速度快,可用于电力系统低碳调度的辅助决策依据,还可用于电力用户低碳生产和生活的引导信息。

Electricity carbon reduction is a key part of achieving national dual carbon goals.The carbon metering method based on carbon flow analysis can better reflect the real carbon emissions of power generation,power grid and users.However,due to the large amount of calculation and time lag of carbon flow analysis method,it is difficult to support low-carbon power dispatching and low-carbon production.To solve the above problems,this paper proposes a neural network based carbon emission factor prediction method for power system nodes,and carries out simulation verification of this method.The results show that this method has high accuracy and fast calculation speed,and can be used as an auxiliary decision-making basis for low-carbon dispatching of power system,and also can be used as guidance information for low-carbon production and life of power users.