Integrating Physics-Informed Neural Networks into Power System Dynamic Simulations
arxiv(2024)
摘要
Time-domain simulations in power systems are crucial for ensuring power
system stability and avoiding critical scenarios that could lead to blackouts.
The proliferation of converter-connected resources, however, adds significant
additional degrees of non-linearity and complexity to these simulations. This
drastically increases the computational time and the number of critical
scenarios to be considered. Physics-Informed Neural Networks (PINN) have been
shown to accelerate these simulations by several orders of magnitude. This
paper introduces the first natural step to remove the barriers for using PINNs
in time-domain simulations: it proposes the first method to integrate PINNs in
conventional numerical solvers. Integrating PINNs into conventional solvers
unlocks a wide range of opportunities. First, PINNs can substantially
accelerate simulation time, second, the modeling of components with PINNs
allows new ways to reduce privacy concerns when sharing models, and last,
enhance the applicability of PINN-based surrogate modeling. We demonstrate the
training, integration, and simulation framework for several combinations of
PINNs and numerical solution methods, using the IEEE 9-bus system.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要