High-Impedance Fault Detection Using Adaptive Soft Denoising
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-impedance faults in electric power distribution circuits are difficult to detect in real-time due to their dynamic and random nature, often resulting in false alarms and challenges in distinguishing between healthy and faulty operating conditions.
Innovation Solution
The implementation of adaptive soft denoising followed by discrete wavelet transform (DWT) for signal processing to detect high-impedance faults, using a two-step method that filters input signals to reduce false alarms and accurately identify fault conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional detection methods are used for high-impedance faults, then the detection process is simple, but the false alarm rate increases and reliability decreases
Solution Approach 1:
The signal processing is divided into multiple stages: initial signal acquisition, wavelet transform decomposition into approximation and detail coefficients, thresholding of detail coefficients, and reconstruction. This segmentation allows complex fault detection to be broken down into manageable processing steps, improving reliability without overwhelming system complexity
Solution Approach 2:
The wavelet transform and thresholding operations are performed as preliminary processing steps before final fault determination. By pre-processing the signal to extract relevant features and remove noise components beforehand, the system establishes a solid foundation for accurate fault detection, reducing false alarms while maintaining structured complexity
2Reliability
If signal filtering is applied to reduce false alarms, then the false alarm rate decreases, but the processing time increases
Solution Approach 1:
The wavelet transform extracts only the relevant frequency components and transient features from the full signal, separating useful information from noise. By taking out only the essential signal characteristics needed for fault detection rather than processing the entire signal spectrum, the system reduces false alarms while limiting the time penalty to only the necessary processing steps
Solution Approach 2:
The threshold parameter is dynamically adjusted based on signal characteristics to optimize the balance between noise rejection and true fault detection. By changing the threshold parameter adaptively rather than using fixed conservative values, the system achieves reliable false alarm reduction without requiring excessive processing time for multiple threshold checks
3Measurement precision
If wavelet transform is used for signal decomposition, then the detection accuracy improves, but the computational complexity increases
Solution Approach 1:
The wavelet transform decomposes the signal into distinct approximation and detail coefficient components at different scales. This segmentation allows the system to focus computational effort on the detail coefficients which contain the fault information, while the approximation coefficients serve as a simpler reference, thereby improving detection accuracy without requiring full processing of all signal components at equal complexity
Solution Approach 2:
The implementation performs wavelet decomposition to a specific level that provides sufficient accuracy for fault detection without going to maximum possible decomposition depths. By applying partial action - just enough decomposition to extract fault features - the system achieves the necessary measurement precision while avoiding the exponential increase in computational complexity that would result from excessive decomposition levels
Data Source
AI summary
Real-time detection of high-impedance faults in a distribution circuit is described. The real-time detection of high-impedance faults includes two steps. First, adaptive soft denoising is employed to perform a filtering process on a healthy dataset, and to determine a threshold. This reduces the rate of false alarms. Second, faulty datasets are prefiltered via adaptive soft denoising, then the denoised signals are processed via discrete wavelet transform to perform high-impedance fault detection using the threshold.