Non-intrusive load decomposition method based on waveform distortion characteristic time sequence change
Patent Information
- Application Number
- CN202511396854.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-28
- Publication Date
- 2025-10-31
- Estimated Expiration
- 2045-09-28
AI Technical Summary
Existing non-intrusive load decomposition technologies suffer from several problems, including high-power appliances masking the state characteristics of low-power appliances, insufficient feature transferability due to load diversity and similar electrical parameters, small differences in load state characteristics, and poor adaptability to dynamic operating conditions.
The current signal of the power circuit is collected by a high-precision current transformer. Using a synchronous sampling and holding and real-time stream processing architecture, the effective value, maximum value, harmonic amplitude and differential signal characteristics of the current signal subsequence are calculated to form a state feature matrix, eliminate outlier errors, obtain the power load feature vector, and realize load decomposition.
It enables precise load status monitoring of power networks, improves dynamic operating condition adaptability and computing resource efficiency, and solves the problems of insufficient feature transferability and small differences in load status characteristics.
Smart Images

Figure CN120879618A_ABST
Abstract
Citation Information
Patent Citations
Canonical correlation analysis-based non-invasive load identification method
CN107832705A
Steady-state current model-based non-intrusive type load decomposition method
CN108257044A