On Approximating the Dynamic Response of Synchronous Generators via Operator Learning: A Step Towards Building Deep Operator-based Power Grid Simulators
arXiv:2301.12538v2 Announce Type: replace-cross Abstract: This paper develops an Operator Learning framework for approximating the dynamic response of synchronous generators. The framework can be used to (i) build a neural network-based generator model that interacts with a power grid simulator or (ii) shadow the true generator's transient response. First, we develop a data-driven Deep Operator Network (DeepONet) to approximate the infinite-dimensional solution operator of the generators. Then, we design a numerical scheme based on DeepONet that simulates the generator's response over a given
The increasing complexity and demands on modern power grids necessitate more advanced simulation tools, and AI/operator learning offers a novel approach to improving their dynamic response modeling.
This development could significantly enhance the resilience and efficiency of power grids by providing more accurate and faster simulations, which are critical for grid management and stability.
The ability to approximate complex generator dynamics via operator learning could lead to more robust power grid simulators, reducing reliance on traditional, slower modeling techniques.
- · Power grid operators
- · Energy management software developers
- · AI/ML solution providers in energy
- · Smart grid technology companies
- · Developers of legacy power grid simulation software
More accurate and faster dynamic response analysis of synchronous generators will be possible.
Improved grid stability and reliability through better predictive capabilities and preemptive intervention.
Enhanced resilience of critical infrastructure against both operational challenges and potential cyber threats via advanced simulation-based security assessments.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI