Two-stage optimal scheduling of CHP microgrid considering renewable energy correlation and risk control

To address the temporal coupling characteristics inherent in renewable energy generation and the dispatch deviations induced by extreme risk events, this study proposes a two-stage optimal scheduling framework considering renewable energy correlation and risk control. A Clayton copula-based joint distribution model is first constructed using historical wind and solar generation data to accurately characterize their interdependence. Subsequently, stochastic renewable energy scenarios are generated via Markov Chain Monte Carlo (MCMC) sampling, followed by inverse transform sampling to obtain a representative scenario set. Building upon this, a scenario-based stochastic optimization model integrating conditional value-at-risk (CVaR) is formulated for the day-ahead scheduling stage, aiming to minimize total system costs by jointly considering operational expenditures and risk exposure. In the intra-day scheduling stage, a model predictive control (MPC) strategy is employed to dynamically adjust scheduling decisions in response to ultra-short-term forecast deviations, thereby enhancing real-time system adaptability. The proposed method is validated through comprehensive case studies, which demonstrate a reduction in total scheduling cost by 2.80 % and in risk cost by 3.30 % compared to conventional methods, while achieving a renewable utilization rate of 94.79 %. The proposed method enables operators to better balance operational efficiency and risk mitigation.