Gaussian process surrogate model to approximate power grid simulators -- An application to the certification of a congestion management controller

📅 2025-02-28
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
Digitalization of power systems invalidates traditional physical models, and reliance on high-fidelity simulators renders large-scale numerical experiments—such as safety verification—computationally infeasible. Method: We propose an adaptive Gaussian process (GP) surrogate model tailored for non-Gaussian power system simulators. Its core innovation is a residual-driven uncertainty correction mechanism that relaxes GP’s strong assumption of Gaussian output distributions. By jointly modeling the predictive mean and residual uncertainty, the method enhances both approximation accuracy and statistical reliability under non-Gaussian conditions. Contribution/Results: The surrogate enables statistically rigorous safety certification of congestion management controllers, reducing original simulator calls by over 98% while preserving statistical confidence—thereby achieving substantial computational savings.

Technology Category

Application Category

📝 Abstract
With the digitalization of power grids, physical equations become insufficient to describe the network's behavior, and realistic but time-consuming simulators must be used. Numerical experiments, such as safety validation, that involve simulating a large number of scenarios become computationally intractable. A popular solution to reduce the computational burden is to learn a surrogate model of the simulator with Machine Learning (ML) and then conduct the experiment directly on the fast-to-evaluate surrogate model. Among the various ML possibilities for building surrogate models, Gaussian processes (GPs) emerged as a popular solution due to their flexibility, data efficiency, and interpretability. Their probabilistic nature enables them to provide both predictions and uncertainty quantification (UQ). This paper starts with a discussion on the interest of using GPs to approximate power grid simulators and fasten numerical experiments. Such simulators, however, often violate the GP's underlying Gaussian assumption, leading to poor approximations. To address this limitation, an approach that consists in adding an adaptive residual uncertainty term to the UQ is proposed. It enables the GP to remain accurate and reliable despite the simulator's non-Gaussian behaviors. This approach is successfully applied to the certification of the proper functioning of a congestion management controller, with over 98% of simulations avoided.
Problem

Research questions and friction points this paper is trying to address.

Reducing computational burden in power grid simulations
Improving accuracy of Gaussian process surrogate models
Certifying congestion management controller functionality efficiently
Innovation

Methods, ideas, or system contributions that make the work stand out.

Gaussian process surrogate model for power grid simulation
Adaptive residual uncertainty term for non-Gaussian behaviors
Machine learning reduces computational burden in grid certification
🔎 Similar Papers
No similar papers found.
P
Pierre Houdouin
RTE, PARIS, FRANCE
M
Manuel Ruiz
RTE, PARIS, FRANCE
L
Lucas Saludjian
RTE, PARIS, FRANCE