CNN Arc Fault Detection Across Variable Electrical Loads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional arc fault detection methods are inadequate in adapting to diverse load environments and fail to effectively detect series arc faults due to reliance on fixed thresholds, leading to incomplete protection against electrical fires.
Innovation Solution
A method utilizing a convolutional neural network for fault arc signal detection, which involves filtering current signals with band-pass filters, constructing time-frequency eigenvectors, and training a two-dimensional convolutional neural network to accurately identify arc faults by analyzing time and frequency features, enabling online determination and adaptive threshold comparison.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional threshold-based arc detection method is used, then the detection method is simple to implement, but it cannot adapt to different load environments and has low detection accuracy
Solution Approach 1:
The patent transforms the detection approach from using fixed thresholds to using a convolutional neural network that automatically learns optimal detection parameters from training data. The system extracts time-frequency features (time-domain and frequency-domain parameters) and uses the neural network to adaptively determine arc faults based on these parameters, enabling adaptation to different load environments without manual threshold adjustment.
Solution Approach 2:
The patent replaces the conventional mechanical threshold-comparison method with an intelligent convolutional neural network system. The neural network automatically processes time-frequency feature matrices and makes detection decisions, substituting the simple but inflexible mechanical thresholding approach with a more sophisticated adaptive system.
2Measurement precision
If conventional overcurrent protection device is used, then the device structure is simple, but it cannot effectively detect series arc faults with current below protection threshold
Solution Approach 1:
The patent moves beyond single-dimensional current magnitude detection to multi-dimensional time-frequency analysis. By extracting both time-domain features (time dispersion, amplitude dispersion, number of waveforms) and frequency-domain features, and organizing them into a two-dimensional feature matrix, the system detects arc faults in multiple dimensions, enabling detection of series arc faults that conventional single-threshold methods miss.
Solution Approach 2:
The patent introduces a convolutional neural network as an intermediary between raw current signals and detection decisions. The neural network processes the complex time-frequency feature matrix and translates it into reliable arc fault detection results, acting as an intelligent mediator that bridges the gap between simple current measurement and accurate fault detection.
3Reliability
If fixed threshold is used for arc detection, then the detection method is easy to operate, but it fails to adapt to diverse loads and produces inaccurate results
Solution Approach 1:
The patent performs preliminary training of the convolutional neural network using training datasets that represent different load environments. This preliminary action enables the system to learn optimal detection patterns beforehand, so that during actual operation, the pre-trained network can reliably adapt to various loads without requiring manual intervention or threshold adjustment, maintaining both reliability and ease of operation.
Data Source
AI summary
A fault arc signal detection method using a convolutional neural network, comprising: enabling a sampling signal subjected to analog-digital conversion to respectively pass through three different band-pass filters; respectively extracting a time-domain feature and a frequency-domain feature from a half wave output of each filter; constructing a two-dimensional feature matrix by means of extracted time-frequency feature vectors from the output of each filter, and stacking the feature matrices corresponding the outputs of the three filters to construct a three-dimensional matrix for each half wave; and processing a multi-channel feature matrix by using a multi-channel two-dimensional convolutional neural network, and determining, according to the output result of the neural network, whether the half wave is an arc. The detection method based on the convolutional neural network has higher accuracy and reliability in recognizing a fault arc half wave, can implement targeted training for different load conditions, and is self-adaptive.