基于时间序列早期分类算法的励磁涌流识别方法及系统
By using a time-series early classification algorithm and LSTM network for multi-dimensional data analysis, transformer current and voltage waveform features are extracted, solving the accuracy problem of transformer inrush current identification and realizing early identification and reliability improvement of transformer protection devices.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING SIFANG JIBAO ENG TECH
- Filing Date
- 2024-12-31
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies make it difficult to accurately identify transformer inrush currents in advance, leading to malfunctions or failures of differential protection devices, which affects the reliability of transformer operation.
An early classification algorithm based on time series is adopted. By collecting voltage and current waveforms of transformers under different states, feature quantities such as second harmonic ratio, waveform symmetry, time-domain equivalent impedance and phase current change rate are extracted. LSTM network is used for multi-dimensional data analysis and multi-confidence fusion to achieve early identification of excitation inrush current.
It improves the accuracy and reliability of inrush current identification, enabling the identification of inrush current in advance, avoiding malfunctions, and enhancing the reliability of transformer protection devices.
Smart Images
Figure CN119989040B_ABST