{
  "abstract": "Operating modern power grids requires coordinated solutions for optimal dispatch, state estimation, price prediction, and contingency analysis, yet these tasks are usually modeled separately. We present PhysBridge, a response-oriented pretraining framework in which historical windows provide context while solver-generated perturbation responses, active constraints, and local sensitivities define the pretraining targets. PhysBridge combines a temporal-graph encoder with a Primal-Dual Response Bridge (PDR-Bridge) that performs safeguarded local refinement and outputs transferable primal, projection-dual, sensitivity, and health tokens. Across IEEE 14-, 30-, 57-, 118-, and 300-bus systems, the benchmark contains 26.3 million solver-generated perturbation-response pairs; in the 300-bus transfer setting, pretraining uses only the 21.04 million nominal pairs from the 14-, 30-, 57-, and 118-bus source topologies. Across three runs on unseen IEEE 300-bus transfer with 10% labels, PhysBridge obtains a 2.21±0.05% OPF cost gap, 96.3±0.2% feasibility, 1.84±0.05 $/MWh LMP MAE, and 91.6±0.3% contingency/fault event macro F1. Relative to a matched rollout-only model trained on the same source corpus without response targets or bridge refinement, these values correspond to a 50.8% lower OPF gap and a 7.4 percentage-point higher event F1. The bridge requires 12.4 ms per 300-bus forward pass and activates fallback on 4.6% of stressed windows. Comparisons with supervised-only OPF specialists that do not receive matched solver-generated pretraining are interpreted as task-relevant performance under the proposed pretraining regime, not as architecture-only superiority.",
  "abstract_shown": true,
  "affiliations": [
    "State Grid Shandong Electric Power Research Institute, Jinan, Shandong, China.",
    "Shandong Key Laboratory of Energy Industry Internet Big Data Technology, Jinan, Shandong, China.",
    "State Grid Shandong Electric Power Company Linyi Power Supply Company, Linyi, Shandong, China."
  ],
  "article_type": "Research Article",
  "arxiv": null,
  "author_total": 5,
  "authors": [
    {
      "affs": [
        0,
        1
      ],
      "name": "Donglan Liu",
      "orcid": "0000-0003-0166-6077"
    },
    {
      "affs": [
        0,
        1
      ],
      "name": "Shanjie Xu"
    },
    {
      "affs": [
        0,
        1
      ],
      "name": "Yuhui Jin"
    },
    {
      "affs": [
        0,
        1
      ],
      "name": "Honglei Yao"
    },
    {
      "affs": [
        2
      ],
      "name": "Qinguo Liu"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/cjsveh546a1v4.bib",
  "cited_by": 0,
  "doi": "10.1371/journal.pone.0359953",
  "fulltext_url": "https://europepmc.org/articles/PMC13716235",
  "id": "cjsveh546a1v4",
  "issns": [
    "1932-6203"
  ],
  "landing_url": "https://doi.org/10.1371/journal.pone.0359953",
  "language": "eng",
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "heading",
      "Artificial Intelligence"
    ],
    [
      "heading",
      "Behavior"
    ],
    [
      "heading",
      "Biology and Life Sciences"
    ],
    [
      "heading",
      "Cognitive Psychology"
    ],
    [
      "heading",
      "Cognitive Science"
    ],
    [
      "heading",
      "Computer and Information Sciences"
    ],
    [
      "heading",
      "Conditioned Response"
    ],
    [
      "heading",
      "Electrical Engineering"
    ],
    [
      "heading",
      "Electrical Faults"
    ],
    [
      "heading",
      "Energy and Power"
    ],
    [
      "heading",
      "Engineering and Technology"
    ],
    [
      "heading",
      "Forecasting"
    ],
    [
      "heading",
      "Learning"
    ],
    [
      "heading",
      "Learning and Memory"
    ],
    [
      "heading",
      "Machine Learning"
    ],
    [
      "heading",
      "Mathematical and Statistical Techniques"
    ],
    [
      "heading",
      "Mathematics"
    ],
    [
      "heading",
      "Neuroscience"
    ],
    [
      "heading",
      "Optimization"
    ],
    [
      "heading",
      "Physical Sciences"
    ],
    [
      "heading",
      "Power Distribution"
    ],
    [
      "heading",
      "Power Grids"
    ],
    [
      "heading",
      "Psychology"
    ],
    [
      "heading",
      "Research Article"
    ],
    [
      "heading",
      "Research and Analysis Methods"
    ],
    [
      "heading",
      "Social Sciences"
    ],
    [
      "heading",
      "Statistical Methods"
    ],
    [
      "heading",
      "Statistics"
    ],
    [
      "heading",
      "Topology"
    ],
    [
      "openalex",
      "2208|Electric Power System Optimization"
    ],
    [
      "plos",
      "/Biology and life sciences/Neuroscience/Cognitive science/Cognitive psychology/Learning"
    ],
    [
      "plos",
      "/Biology and life sciences/Neuroscience/Learning and memory/Learning"
    ],
    [
      "plos",
      "/Biology and life sciences/Psychology/Behavior/Conditioned response"
    ],
    [
      "plos",
      "/Biology and life sciences/Psychology/Cognitive psychology/Learning"
    ],
    [
      "plos",
      "/Computer and information sciences/Artificial intelligence/Machine learning"
    ],
    [
      "plos",
      "/Engineering and technology/Electrical engineering/Electrical faults"
    ],
    [
      "plos",
      "/Engineering and technology/Energy and power/Power distribution/Power grids"
    ],
    [
      "plos",
      "/Physical sciences/Mathematics/Optimization"
    ],
    [
      "plos",
      "/Physical sciences/Mathematics/Statistics/Statistical methods/Forecasting"
    ],
    [
      "plos",
      "/Physical sciences/Mathematics/Topology"
    ],
    [
      "plos",
      "/Research and analysis methods/Mathematical and statistical techniques/Statistical methods/Forecasting"
    ],
    [
      "plos",
      "/Social sciences/Psychology/Behavior/Conditioned response"
    ],
    [
      "plos",
      "/Social sciences/Psychology/Cognitive psychology/Learning"
    ]
  ],
  "openalex": "W7224098711",
  "pdf_url": "https://europepmc.org/articles/PMC13716235?pdf=render",
  "pmcid": "PMC13716235",
  "pmid": "42949919",
  "published": "2026-10-28",
  "publisher": "Public Library of Science",
  "ris": "https://research.draconitic.com/paper/cjsveh546a1v4.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:17:17Z",
      "native_id": "10.1371/journal.pone.0359953",
      "source": "plos",
      "url": "https://doi.org/10.1371/journal.pone.0359953"
    },
    {
      "first_seen": "2026-10-31T00:20:33Z",
      "native_id": "MED:42949919",
      "source": "europepmc",
      "url": "https://europepmc.org/article/MED/42949919"
    },
    {
      "first_seen": "2026-11-01T12:23:16Z",
      "native_id": "PMC13716235",
      "source": "pmc",
      "url": "https://doi.org/10.1371/journal.pone.0359953"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "plos:neuroscience",
      "subfield": "neuroscience"
    },
    {
      "field": "mathematics",
      "origin": "plos:topology",
      "subfield": "geometry-topology"
    },
    {
      "field": "mathematics",
      "origin": "plos:statistical methods",
      "subfield": "statistics"
    }
  ],
  "title": "PhysBridge: Response-supervised grid pretraining with physics-constrained local correction for transferable power-system intelligence",
  "title_text": "PhysBridge: Response-supervised grid pretraining with physics-constrained local correction for transferable power-system intelligence",
  "type": "peer-reviewed",
  "updated_at": "2026-11-01T12:23:16Z",
  "url": "https://research.draconitic.com/paper/cjsveh546a1v4",
  "venue": "PLOS One",
  "versions": []
}