基于知识蒸馏算法的非侵入式负荷监测方法及系统
By building a lightweight neural network model based on the knowledge distillation algorithm on edge devices, the deployment problem of non-intrusive load monitoring on the user side is solved, achieving efficient monitoring and identification of electrical equipment, and reducing communication costs and privacy risks.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- STATE GRID TIANJIN ELECTRIC POWER COMPANY
- Filing Date
- 2024-12-24
- Publication Date
- 2026-07-17
AI Technical Summary
Existing non-intrusive load monitoring methods require transmitting edge data to the cloud for training, resulting in high communication costs and privacy and security risks, making them inefficient to deploy on user-side edge devices.
A lightweight neural network model based on the knowledge distillation algorithm is adopted. By constructing teacher and student neural networks and combining data generation and knowledge distillation algorithms, the model is trained and validated on edge devices, reducing model parameters and computational load, and realizing the monitoring and identification of load devices.
The ability to quickly monitor and identify electrical devices on edge devices reduces the number of parameters and computational complexity in model training, decreases communication costs, and improves privacy and security.
Smart Images
Figure CN119830999B_ABST