{
  "$type": "site.standard.document",
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreibjtjabuz6y5wkh7dhknqjmf3oexyd2jdmj2qcasd3zb6ef6l4ndu"
    },
    "mimeType": "image/png",
    "size": 123715
  },
  "description": "The present invention discloses an XGBoost-based lithium-ion battery thermal runaway acoustic source localization error compensation method. Acoustic signals in an energy storage cabin are collected using a fixed microphone array, including time delay data, cross power spectrum data, and acoustic…",
  "path": "/patents/1423536",
  "publishedAt": "2026-06-18T00:00:00.000Z",
  "site": "at://did:plc:oql6ds5vnff4ugar6rruliwd/site.standard.publication/3mn3ohu7oxx5w",
  "tags": [
    "H01M10/486",
    "Nanjing Tech University"
  ],
  "textContent": "The present invention discloses an XGBoost-based lithium-ion battery thermal runaway acoustic source localization error compensation method. Acoustic signals in an energy storage cabin are collected using a fixed microphone array, including time delay data, cross power spectrum data, and acoustic source localization coordinates calculated by a conventional geometric method. The localization error is calculated by comparing the geometric localization coordinates with actual acoustic source position coordinates, and a feature vector is constructed based on the extracted acoustic features. These feature vectors are used to train an XGBoost model. In practical applications, real-time time delay data, cross power spectrum data, and geometric localization coordinates are input into the trained model, and the predicted error value is used to correct the geometric acoustic source localization results, thereby obtaining a more accurate acoustic source position for a thermal runaway lithium-ion battery.",
  "title": "XGBoost-Based Error Compensation Method for Acoustic Source Localization of Lithium-Ion Battery Thermal Runaway"
}