ArcNet-Lite CNN for Real-Time Series Arc Fault Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing arc fault detection systems face challenges in accurately and efficiently detecting series arc faults in real-time, particularly due to the limitations of conventional algorithms and the need for reduced computational complexity in edge devices.
Innovation Solution
The development of a lightweight deep neural network model, ArcNet-Lite, which employs a convolutional neural network architecture optimized for embedded hardware and utilizes a teacher-student knowledge distillation technique to achieve high accuracy in arc fault detection while minimizing computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional arc fault detection algorithms are used, then detection capability is provided, but detection accuracy for series arc faults is insufficient and computational complexity is high
Solution Approach 1:
The patent replaces conventional signal processing algorithms with a neural network-based system that uses electro-optical sensing. The neural network processes optical signals from arc detection circuits rather than traditional electrical current analysis, substituting one detection mechanism for another more effective one. This achieves superior detection accuracy while the embedded implementation keeps computational complexity manageable for edge devices.
Solution Approach 2:
The system combines multiple detection approaches into a composite detection system: optical arc detection circuits are integrated with electrical current sensing, and the neural network combines features from multiple input sources. This composite approach leverages the strengths of different detection methods to achieve high accuracy in series arc fault detection.
2Speed
If real-time detection is implemented, then response speed is improved, but computational resources required increase
Solution Approach 1:
The patent segments the detection system into distinct functional modules: optical arc detection circuits that capture arc events, separate neural network processing units that analyze the signals, and control circuits that implement protection actions. This segmentation allows real-time processing by distributing computational tasks across specialized components, reducing the resource burden on any single element while maintaining fast response.
Solution Approach 2:
The neural network acts as an intermediary between the raw optical/electrical signals and the final detection decision. It processes intermediate representations of the arc fault conditions, transforming complex signal patterns into actionable detection outcomes. This intermediary processing layer enables real-time analysis by breaking down the computational task into manageable transformation steps.
Data Source
AI summary
An apparatus obtains an input signal representative of a current passing through a first circuit and applies the input signal to one or more input nodes of a second circuit configured according to a convolutional neural network model. The apparatus drives one or more output nodes of the second circuit according to a detection by the convolutional neural network model of an arc fault in the current passing through the first circuit. The one or more output nodes are driven to either a first value indicative of an absence of the detection of the arc fault in the current passing through the first circuit, or a second value indicative of the detection of the arc fault in the current passing through the first circuit. A switch through which the current passes remains closed in response to the first value and opens in response to the second value.