Flexible Scheduling Strategy Based on Situation Awareness Technology under Extreme Weather Events

With the structural morphology evolution of the new power system paradigm, the emergence of random extreme weather events may lead to scenarios of low renewable energy power output and a sharp increase in load demand for a single day or multiple consecutive days, which seriously affect the power supply guarantee of the system. Perceiving the system’s extreme operation scenarios in advance and developing flexible scheduling schemes will help to enhance the system’s resilience. Firstly, based on the framework of the situation awareness technology, this paper proposes a physical-data co-driven model, including the sequential operation optimization model and convolutional neural network model, to explore the mapping relationship between the meteorological perception data and the system’s operation situation labels, realizing the advanced awareness of extreme operation risk scenarios. Secondly, a flexible scheduling strategy with variable balance periods is developed to optimize the system’s unit commitment and energy storage level, improving the system’s overall adaptation resilience and economic benefits. The case study demonstrates the superiority of the proposed model and strategy.