Non-intrusive Load Monitoring via V-I Trajectory RGB Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current non-intrusive load monitoring methods fail to fully utilize steady-state V-I trajectory features and power features, leading to incomplete energy monitoring and inefficient energy savings.
Innovation Solution
A non-intrusive load monitoring method that collects real-time voltage, current, and power data, uses a bilateral sliding window algorithm to detect switching events, and converts V-I trajectory into an RGB image for analysis by a convolutional neural network, enabling the full utilization of both V-I trajectory and power features for load monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional NILM methods use only power features or only V-I trajectory features, then the monitoring cost remains low, but the monitoring accuracy is insufficient
Solution Approach 1:
The patent merges V-I trajectory features and power features into a unified monitoring framework. The V-I trajectory is converted into an RGB image where power information is encoded as pixel values, allowing the convolutional neural network to simultaneously process both trajectory shape information and power magnitude information, thereby achieving comprehensive feature utilization and improved monitoring accuracy
Solution Approach 2:
The patent transforms the V-I trajectory from a traditional 2D plot into a 3D RGB image structure where the third dimension (color channel) represents power information. This dimensional transformation allows the neural network to capture both spatial trajectory patterns and power magnitude simultaneously, resolving the information loss caused by using only single-type features
2Measurement precision
If cloud servers are used for data processing, then complex computations can be performed, but the system cost and complexity increase
Solution Approach 1:
The patent replaces traditional server-based cloud processing with an embedded monitoring module that runs directly on the electricity meter. By using a convolutional neural network optimized for embedded devices, the system performs complex data processing locally without requiring external cloud servers, thereby reducing system complexity and cost while maintaining processing capability
Solution Approach 2:
The embedded monitoring module on the electricity meter performs self-processing of voltage, current, and power data. The module independently executes the V-I trajectory conversion and neural network analysis without relying on external cloud infrastructure, enabling the system to serve itself and eliminating the need for complex server-based processing
Data Source
AI summary
A non-intrusive load monitoring method based on V-I trajectory and neural network includes: collecting the household voltage, current and active power data in real time; determining whether there is a switching event and whether the load operating state has reached a steady state through the change of the active power; obtaining the voltage, current and power data of the load, converting the V-I trajectory into RGB color image containing the phase difference between the voltage and current, power and other information. After obtaining the RGB color image, performing normalization processing and performing load monitoring through pre-trained convolutional neural network. The present disclosure fully extracts the steady-state feature of the load through the convolutional neural network, and the neural network model can directly run on an embedded device, and does not need to rely on the computing support of a server.