A method for fault location prediction of traveling wave transmission characteristics of a multi-type cable architecture

By optimizing sensor placement and establishing a dynamic reconstruction model in various cable architectures, and combining feature fusion with hyperbolic wavelet transform and attention mechanism, the problems of blind sensor placement and incomplete feature extraction are solved, enabling accurate location and timely early warning of cable faults, and improving the accuracy and reliability of fault prediction.

CN119959681BActive Publication Date: 2026-06-19GUANGZHOU POWER SUPPLY BUREAU GUANGDONG POWER GRID CO LTD

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
GUANGZHOU POWER SUPPLY BUREAU GUANGDONG POWER GRID CO LTD
Filing Date
2025-01-14
Publication Date
2026-06-19

Smart Images

  • Figure CN119959681B_ABST
    Figure CN119959681B_ABST
Patent Text Reader

Abstract

This invention discloses a fault location and prediction method for traveling wave transmission characteristics in multi-type cable architectures, belonging to the field of fault location technology. The method includes: real-time monitoring of multi-type cable networks to acquire state data streams containing traveling wave characteristics; establishing a dynamic reconstruction model of cable traveling wave transmission characteristics; performing time-frequency domain joint analysis on the state data streams to extract fault feature vector sets; constructing a fault mapping network based on the fault feature vector sets; the fault mapping network includes a fault type feature matrix and a location feature matrix; performing collaborative operations on the fault type feature matrix and the location feature matrix to complete the joint prediction of fault type and location, and outputting fault early warning information. This invention achieves accurate fault location and timely early warning in multi-type cable architectures, improving the accuracy and reliability of fault prediction, and providing effective technical support for preventive maintenance of cable systems.
Need to check novelty before this filing date? Find Prior Art