A method for constructing a time sequence attention network model for early warning of piglet diarrhea disease

CN120809251BActive Publication Date: 2026-04-21LONGYAN UNIV
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
LONGYAN UNIV
Filing Date
2025-09-16
Publication Date
2026-04-21

AI Technical Summary

Technical Problem

Existing early warning models for swine diarrhea mostly rely on data from a single sensor, ignoring the spatiotemporal correlation between environmental factors and physiological indicators. This results in weak early signals being masked by noise, limiting the effectiveness of early warning.

Method used

A temporal attention network model for early warning of swine diarrhea was constructed. By acquiring physiological indicators and environmental factor data, preprocessing and feature extraction were performed. An initial feature vector was generated using a deep learning time series modeling module. Temporal weight labeling and cross-modal temporal correlation analysis were performed to generate a multimodal feature matrix and map it with a health status knowledge graph to generate early warning entries.

Benefits of technology

It significantly improves the accuracy and reliability of early warning of swine diarrhea, effectively capturing early weak signals and providing scientific health management support.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120809251B_ABST
    Figure CN120809251B_ABST
Patent Text Reader

Abstract

This invention relates to the field of swine health monitoring technology and discloses a method for constructing a temporal attention network model for early warning of swine diarrhea. The method includes: acquiring and preprocessing swine physiological indicators and environmental factor data to generate standardized sequences; extracting features through deep learning to generate an initial feature vector set; labeling temporal weights to determine timestamp tags and dynamic weight coefficients; performing cross-modal temporal correlation analysis to generate the direction and strength of inter-modal correlations; collecting multimodal data to construct a feature matrix and performing temporal alignment; and mapping the aligned feature matrix to a health status knowledge graph to generate warning entries. This invention combines deep learning and temporal attention mechanisms to uncover deep correlations in multimodal data, improving the accuracy and efficiency of early warning and providing reliable technical support for swine health monitoring.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Pig behavior and health method based on deep learning and physiological parameter dynamic monitoring

    CN118986303A

  • Livestock breeding state monitoring method and system

    CN119558336A

  • Jinhua pork quality improvement method and system based on data mining

    CN120634041A