Arc Fault Detection Using Signal Processing and Higher-Order Measures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current arc fault detection systems in complex electrical environments, such as aerospace and residential settings, fail to reliably detect series arc faults due to their low energy levels, often causing malfunctions and fires, and lack universal methods to differentiate between arc faults and normal transient signals.
Innovation Solution
A generalized arc fault detection system that inputs AC or DC current signals, extracts fundamental components, monitors amplitude variations, and applies higher-order measures to detect non-stationary changes, allowing for real-time identification of series and parallel arc faults with improved noise immunity and reduced nuisance trips.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional circuit breakers are used to detect arc faults, then over-current and overload conditions can be detected, but series arc faults with low energy levels cannot be detected
Solution Approach 1:
The patent transforms the current signal through multiple processing stages including FFT transformation, filtering to extract specific frequency components, and statistical analysis. By changing the parameters of signal processing (frequency domain transformation, selective filtering, statistical moment calculation), the system can detect subtle arc fault signatures that conventional breakers miss, while maintaining manageable complexity through systematic signal transformation
Solution Approach 2:
The patent replaces conventional mechanical/electromagnetic detection mechanisms with digital signal processing methods. Instead of relying on simple thermal or magnetic effects in circuit breakers, the system uses computational approaches (FFT, statistical analysis, pattern recognition) to detect arc faults, enabling detection of low-energy series arcs through intelligent algorithms rather than physical thresholds
2Measurement precision
If filtering and threshold detection methods are used to distinguish arc faults from normal signals, then arc detection can be achieved, but nuisance trips occur frequently due to inability to differentiate from transient signals
Solution Approach 1:
The patent performs preliminary signal processing including FFT transformation and filtering before detection. By pre-processing the signal to extract relevant frequency components and remove noise, the system prepares the data in advance for more accurate detection, reducing false alarms caused by transient signals that don't exhibit arc-specific patterns in the processed domain
Solution Approach 2:
The system uses statistical analysis and pattern recognition that inherently provide feedback mechanisms. By analyzing multiple signal characteristics (frequency content, statistical moments, temporal patterns) and comparing against learned or predefined arc signatures, the system can distinguish true arc faults from normal transients, improving both detection accuracy and operational stability
3Measurement precision
If detection sensitivity is increased to detect low energy series arc faults, then detection capability improves, but false alarms from normal transients increase
Solution Approach 1:
The patent moves the detection problem from the time domain to the frequency domain using FFT transformation. By analyzing signals in the frequency domain and examining statistical moments, the system creates additional dimensional information that helps distinguish arc faults from transients. This dimensional transformation enables sensitive detection of low-energy arcs while maintaining immunity to false alarms through multi-dimensional signal characterization
Data Source
AI summary
A method and an apparatus detect series and/or parallel arc faults in AC and DC systems. The method according to one embodiment inputs an AC current signal; extracts a fundamental component of the AC current signal and monitoring an amplitude variation profile for the fundamental component, thereby generating a first arc fault detection measure; detects non-stationary changes in the AC current signal applying at least one measure of order higher than one, thereby generating a second arc fault detection measure; and determines whether an arc fault exists based on the first arc fault detection measure and the second arc fault detection measure.