A load sequence scenario generation method based on a dual-layer Markov chain Monte Carlo model

Generating a large number of representative load sequence scenarios is crucial for evaluating the supply-demand balance, economic efficiency, and reliability of power systems, which have a high penetration rate of new energy. Current scenario generation methods have certain limitations, such as the need for extensive historical data and the inability to reflect the relatively fixed patterns of load sequences over the medium to long term. To address these issues, this paper proposes a load sequence scenario generation method according to a dual-layer Markov chain Monte Carlo (MCMC) model. The outer MCMC model captures the probabilistic characteristics of day-type transitions within a month, while the inner MCMC model models the state transition probabilities of the load sequences. Through optimization techniques, the daily load sequences are generated. Case study analysis shows that the statistical characteristics of the scenarios generated by this method are highly similar to the original scenarios and better than the scenarios generated without considering the outer MCMC, providing diverse scenario support for the evaluation of power systems connected to the high proportion of new energy.