Multi-timescale optimization control method for low-carbon operation of power grid under carbon emission constraints

Abstract

Relying on fixed scheduling rules to allocate power plant output can only barely maintain a basic balance between electricity supply and demand, resulting in inefficient grid control. Therefore, this study investigates multi-timescale optimization control methods for low-carbon grid operation under carbon emission constraints. A multi-scale optimization model was constructed, incorporating a multi-objective function that balances economic, low-carbon, and safety objectives, along with a four-category constraint system that includes carbon emissions. A three-tiered collaborative control strategy—comprising day-ahead, day-ahead-to-intraday, and real-time control—was adopted. Day-ahead scheduling provides baseline instructions; day-ahead-to-intraday rolling adjustments correct prediction errors; and real-time control smooths out second-level fluctuations while ensuring that instantaneous carbon emissions do not exceed limits. This multi-scale, bidirectional collaborative mechanism achieves optimized control. Experiments demonstrate that this method reduces total carbon emissions to as low as approximately 0.18 million tons per hour during peak renewable energy generation periods. Under load conditions of 5,000 MW and 7,000 MW, the renewable energy integration rates reach 98% and 99%, respectively, with an average control response time of 2.5 seconds, effectively enhancing the efficiency of power grid control.