Multi-timescale Optimal Scheduling for Carbon-oriented Microgrids Based on an Improved Grey Wolf Optimization Algorithm

With the integration of a high proportion of renewable energy, microgrid operation faces multiple challenges, including increased source-load uncertainty and stricter carbon emission reduction constraints. To address the difficulty of traditional scheduling methods in balancing economy, low-carbon performance, and operational stability, this paper proposes a carbon-oriented multi-timescale optimal scheduling method for microgrids based on an improved grey wolf optimization (GWO) algorithm. First, a carbon-oriented demand response model is developed by integrating low-carbon time-of-use tariffs and a carbon-exceedance reward and penalty mechanism. This model guides load transfer to low-carbon periods. Second, a robust optimization model is established in the day-ahead stage considering wind power, photovoltaic power, and load uncertainties. An improved GWO algorithm based on the similarity principle is introduced to solve the model. Furthermore, model predictive control is adopted in the intraday stage to realize rolling optimization and dynamically correct the day-ahead schedule. The results show that the proposed method effectively achieves peak shaving and valley filling, reduces carbon emissions, and improves the overall operating cost and system stability under uncertainty.