Graph Neural Solver for Power Systems

Source

• Raw Markdown: graph-neural-solver-power-systems-2019
• Rendered / retrieved PDF: paper_graph-neural-solver-power-systems-2019.pdf
• External source: https://hal.science/hal-02175989
• Official L2RPN reference list: https://l2rpn.chalearn.org/papers-references

Publication And Credibility

• Paper date: 2019-07-14.
• Venue/status: IJCNN 2019; HAL PDF retrieved; DOI 10.1109/IJCNN.2019.8851855.
• Credibility: Peer-reviewed IJCNN paper with HAL PDF retrieved locally. Older than one year; use as graph-solver lineage.

Core Claim

The paper proposes a graph neural load-flow surrogate that uses power-grid topology and injection structure to predict line flows, with experiments showing zero-shot generalization from 30-node random grids to smaller and larger random grids.

L2RPN / Grid2Op Notes

It is an earlier graph-structure source for power systems and helps motivate why topology-aware encoders matter for L2RPN-style environments.

Use this as simulator/surrogate evidence, not as policy evidence: the paper does not provide action/reward trajectories, and its reported transfer uses artificial random grids with equal line characteristics.

Action-Time-Series Notes

This source is useful when Grid2Op is treated as an action-conditioned graph time-series environment:

power-grid observations + topology / redispatch / storage control input + scenario context
  -> next grid observations + safety/cost outcome

The terminology distinction matters. Topology changes, redispatching, curtailment, and storage commands are actions or control inputs when an agent chooses them. Line failures, maintenance outages, weather-driven renewable shifts, and demand variation are events or exogenous variables unless they are deliberately controlled by the experimenter.

Foundation TSFM Relevance

Agenda slot	Verdict	Evidence	Missing pieces
Causal structure, counterfactuals, and control	adjacent	Useful background for graph time series and topology-aware TSFM encoders.	A graph solver is not a policy and does not define logged action/reward trajectories.
Context interface: topology and channel context	partially closes	Power-grid state is naturally graph-structured and tied to physical assets, limits, and scenario metadata.	Needs a reusable schema that a general TSFM can consume across grids and non-grid operational systems.
Benchmark level	adjacent	L2RPN/Grid2Op provides simulator-backed trajectories with explicit controls and outcomes.	TSFM-ready comparisons require pinned environment versions, action sets, reward definitions, and train/test scenario splits.

Links Into The Wiki

• Grid2Op
• Neural Networks for Power Flow: Graph Neural Solver
• Action-Conditioned Time-Series Datasets
• Time-Series Benchmark Hygiene
• World Models
• Foundation Time-Series Model Research Agenda
• Lineage: Fast Power system security analysis with Guided Dropout, Anticipating contingengies in power grids using fast neural net screening, and Optimization of computational budget for power system risk assessment.