Soft Fault Detection in Cables via Data Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing reflectometry methods struggle to accurately detect and locate soft faults in cables due to high rates of false detections, as they amplify both fault signatures and measurement noise, leading to ambiguity in diagnosis.
Innovation Solution
A data fusion method is applied to combine the results from multiple independent fault detection techniques, involving the acquisition of a time reflectogram, difference calculation with a baseline, application of independent transformations, conversion to fault probability, and iterative data fusion to produce a unified probability of defect occurrence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If post-processing is applied to amplify fault signatures in reflectograms, then detection sensitivity for soft faults is improved, but false detection rate increases due to noise amplification
Solution Approach 1:
The patent segments the detection process into multiple independent transformation methods applied separately to the reflectogram. Each method processes the signal differently, and their results are combined through data fusion. This segmentation allows each method to focus on specific fault characteristics while avoiding the amplification of noise that plagues single-method approaches.
Solution Approach 2:
The patent merges results from multiple independent transformation methods through data fusion to produce a unified diagnosis. By combining the strengths of different methods and cross-validating their findings, the system achieves both high detection sensitivity and low false detection rates, resolving the contradiction between amplifying fault signatures and avoiding noise amplification.
2Measurement precision
If multiple independent transformation methods are applied to the reflectogram, then detection accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent implements a universal data fusion framework that can accommodate multiple different transformation methods. This multi-functional approach allows the same fusion mechanism to process results from various independent methods, managing complexity through a standardized integration layer while maintaining high detection accuracy.
Solution Approach 2:
The data fusion process incorporates feedback mechanisms where results from multiple transformations are cross-validated and iteratively combined. This feedback loop refines the detection accuracy by comparing and reconciling findings from different methods, while the systematic feedback structure actually reduces overall processing complexity by providing a clear integration pathway.
3Reliability
If data fusion is applied to combine multiple detection results, then false detection rate is reduced, but computational requirements increase
Solution Approach 1:
The patent applies partial data fusion by selectively combining results from multiple transformation methods only when necessary for ambiguous cases. This partial action approach reduces computational requirements compared to full fusion of all possible method combinations, while still achieving significant reduction in false detection rates for the most critical cases.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly reduces false detection rates by isolating and amplifying only the signatures of soft faults, providing a more reliable and accurate diagnosis of cable conditions.
Implementation Method 1
an electrical signal, the probe signal or reference signal, which is most often high-frequency or broadband, is injected at one or more points in the cable being tested. The signal propagates through the cable or network and reflects some of its energy back when it encounters an electrical discontinuity.
Data Source
Figure 1
Figure 2
Figure 3~5
AI summary
Method for detecting soft faults in a transmission line, comprising the following steps: - acquiring (401) a measurement, termed a time-domain reflectogram, of a characteristic signal of the reflection of a reference signal previously injected in the line; - determining (402) the difference between said time-domain reflectogram and a time-domain reflectogram measured earlier for the same line or an identical line with similar characteristics, in order to obtain a corrected time-domain reflectogram; - applying (4031, 403i, 403N) a plurality of independent transformations to the corrected time-domain reflectogram, in order to obtain a plurality of independent transformed reflectograms; - converting (404, 405, 406, 407) the transformed reflectograms into a plurality of mutually-independent probabilities of occurrence of a fault, - applying (408) a data fusion method to the probabilities of occurrence of a fault in order to deduce a unified value of the probability of occurrence of a fault.