Calculating Probabilistic Carbon Emission Flow: An Adaptive Regression-Based Framework
Mingchen Ma, Yaowang Li, Ershun Du, Haiyang Jiang, Ning Zhang, Wei Wang, Min Wang
Abstract
Carbon intensities are beginning to be used as incentives for consumer-driven carbon reduction. Guided by time-varying carbon intensities, consumers can schedule loads in priority order to use more electricity during low-carbon periods. However, carbon intensities cannot be perfectly forecasted in advance due to the fluctuation of loads and renewable energy. The probabilistic distribution of carbon intensities can be calculated by the grid operator and used as a reference for power consumers. This paper presents a probabilistic carbon emission flow model to calculate the distribution of carbon intensities for consumers. An adaptive regression-based calculation framework combined with a carbon pattern dictionary technique is proposed to handle the high calculation complexity caused by the nonlinearity of the carbon emission flow model. The simulation results from the case studies demonstrate the accuracy and efficiency of our proposed approach.