Arcing Event Detection Using Waveform Shape Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional arcing-event-detection systems in electrical power systems struggle to detect low-magnitude currents and cannot utilize multiple properties of the electrical power system directly, leading to missed hazardous conditions and early stages of system failures.
Innovation Solution
A system that includes a data acquisition unit monitoring signals indicative of periodic properties like current, with a computing device removing steady-state load components and identifying arcing events through bursts with sinusoidal shapes or precipitous decreases, allowing for the analysis of both current and voltage waveforms to determine the presence and cause of arcing events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional overcurrent protection devices are used, then they can detect persistent overcurrent events, but they cannot detect temporary arcing events due to time delays and impedance limitations
Solution Approach 1:
The system performs preliminary analysis of current waveform characteristics before final detection is made. By examining the temporal behavior and shape of current bursts in advance, the system can identify arcing events before they persist long enough to trigger conventional overcurrent devices, effectively acting ahead of time to detect temporary events.
Solution Approach 2:
The system dynamically adapts its detection criteria based on the temporal behavior and characteristics of current bursts. Rather than using fixed thresholds, it analyzes the dynamic shape, duration, and magnitude variations of current waves to distinguish arcing events from normal load variations, enabling detection of temporary events without fixed time delays.
2Measurement precision
If conventional arcing-event-detection systems are used, then they can detect arcing events with sufficient current magnitude, but they cannot detect low-magnitude currents from high-impedance arcing events
Solution Approach 1:
The system segments the current signal into distinct temporal portions, separating the normal load current component from the arcing event component. By analyzing only the burst portions that deviate from the steady-state load pattern, the system can detect low-magnitude arcing currents that would be lost in the noise of normal operational current.
Solution Approach 2:
The system extracts and isolates the arcing event component from the total current signal by removing the steady-state load current. This extraction process allows the system to focus analysis on only the transient portions caused by arcing, enabling detection of low-magnitude currents that are obscured when combined with normal load current.
3Device complexity
If conventional arcing-detection systems analyze single property parameters, then the detection method is simple, but they cannot utilize multiple properties of the electrical power system to improve detection accuracy
Solution Approach 1:
The system merges the analysis of multiple electrical properties (current magnitude, temporal behavior, waveform shape, and duration) into a unified detection approach. By combining these different aspects of electrical behavior, the system achieves higher detection accuracy for arcing events while maintaining manageable complexity through integrated analysis rather than separate detection systems.
Data Source
AI summary
A system for detecting electrical arcing on an electrical power system includes: a) a data acquisition unit that is electrically connected to an electrical power system, wherein the data acquisition unit is configured to monitor signals indicative of a first periodic property of the electrical power system, wherein the signal includes a normal load component; and b) a computing device operably connected to the data acquisition unit. The computing device is programmed to: i) obtain first data from the data acquisition unit indicative of the temporal behavior of the first signal; ii) remove the normal load component from the first data; and iii) determine that an arcing event is present on the electrical power system when the at least one burst within the first data presents a generally sinusoidal shape which includes generally flat regions which are present where the generally sinusoidal shape crosses over a zero-magnitude line.