Low Carbon Oriented Multi-Stage Energy Management of Integrated Energy Systems Considering Spatio-Temporal Correlation Characteristics and Operational Uncertainties

Wanbing Zhao,Weiguang Chang,Qiang Yang

DOI: https://doi.org/10.2139/ssrn.4678219

2023-01-01

Abstract:The integrated energy system (IES) that includes renewable distributed generation (RDG) is considered an efficient system to support low-carbon energy supply, offering economic and environmental benefits. However, the operational uncertainties introduced by intermittent distributed renewable generation and different forms of demands bring technical challenges to the energy management of IES. This paper proposes a multi-stage and multi-timescale low-carbon energy management framework (EMF) incorporating the spatio-temporal correlations and uncertainties of RDGs and demands. The proposed solution encompasses a day-ahead stage and an intra-day stage based on the generation and demand prediction using a spatio-temporal feature graph convolutional network (STFGCN). The day-ahead scheduling model aims to minimize pre-economic costs and carbon emission costs considering a ladder-type carbon trading mechanism (CTM) to achieve low-carbon economic scheduling. The model predictive control (MPC) based intra-day dispatch model aims to minimize economic costs and the penalties for intra-day dispatch deviations through rolling optimization. The proposed method is extensively assessed through experiments against the benchmark solutions, and the numerical results confirmed its effectiveness in mitigating the impact of operational uncertainties with reduced operational cost and carbon emissions.