Economic and stable scheduling strategy for electricity-thermal-hydrogen-gas integrated energy systems based on online MPC under dynamic supply-demand interaction

Predictive-based decision optimization methods have gained traction in integrated energy system operations, particularly when historical source-load data is accessible and uncertainties exist. However, these methods typically isolate prediction from optimization, ignoring current impacts on future states and prioritizing economic benefits, causing load fluctuations and supply volatility that restrict real-time decision-making. To address this challenge, this study proposes a real-time optimization method utilizing online model predictive control (MPC) for a system integrating electricity, thermal energy, hydrogen, and gas. A novel evaluation framework is developed to balance economic benefits with operational stability, accompanied by a dynamic adjustment strategy for control and prediction horizons. Case studies demonstrate the outstanding performance of the online MPC method and illustrate the differences from traditional predictive-based methods. Specifically, our method increases economic benefits by 2.96 %. Under renewable energy disturbances, the proposed method improves energy supply reliability by 6.58 % and 7.76 % compared to traditional methods #1 and #2, respectively. The study underscores how multi-energy and multi-market interactions affect benefits, with halting hydrogen conversion reducing economic gains by 0.61 % and stopping electric vehicle discharging cutting them by 15.97 %. The results highlight the method's effective balance between economy and stability in dynamic supply-demand scenarios, which is promising for system optimization.