关键词: 综合能源系统(IES), 风力发电预测, 碳捕集, 氢能多元结构, 低碳经济调度, 递减时序卷积网络(DTCN), 快速傅里叶变换(FFT)

Abstract: [Objectives] The uncertainty of wind power generation output and the complexity of diverse energy chains pose significant challenges to the low-carbon operation of integrated energy systems (IES). Therefore, this paper proposes a low-carbon economic dispatch model for IES based on DTCN-FFT wind power forecasting. [Methods] First, a decreasing temporal convolutional network (DTCN) combined with a fast Fourier transform (FFT) prediction model is employed to enhance the accuracy of wind power forecasting. Second, lean-rich storage tanks are integrated into the carbon capture system to strengthen its resilience against wind power fluctuations. Third, a diversified hydrogen energy structure encompassing electrolyzer hydrogen production, natural gas hydrogen production, hydrogen fuel cells, and hydrogen storage tanks was constructed to maximize synergies between hydrogen utilization and the carbon capture system. Finally, the model was solved by minimizing the sum of carbon trading costs, coal-fired power and gas procurement costs, equipment maintenance costs, and prediction error compensation costs. [Results] The proposed model reduced the total cost of the IES by 26.62% and decreased carbon emissions by 51.23%. [Conclusions] This model enhances wind power utilization, mitigates hydrogen energy losses, and improves the IES in terms of both low-carbon performance and economic efficiency.

Key words: integrated energy system (IES), wind power forecasting, Carbon capture, Hydrogen multi-structure, Low-carbon economic dispatch, decreasing temporal convolutional network (DTCN), fast fourier transform (FFT)