Partial Discharge Triggering for Transformer Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for locating partial discharge sources in electrical transformers face challenges due to high magnitude noise, high repetition rate noise signals, and parallel activities, which complicate signal acquisition and averaging, especially in real-life transformer conditions with non-ideal configurations such as tank walls and complex signal paths.
Innovation Solution
A method that generates a trigger signal only when the phase angle and amplitude of the partial discharge pulse fall within predefined trigger conditions, allowing for focused measurement on a single source and using multi-dimensional trigger conditions including phase, amplitude, time, frequency, and acoustic emission to assign measurement samples to specific clusters, thereby improving signal acquisition and noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional trigger methods are used for signal acquisition, then all partial discharge activities are recorded, but noise signals and parallel activities complicate the measurement and reduce precision
Solution Approach 1:
The patent applies parameter changes by using multi-dimensional trigger conditions that evaluate multiple parameters simultaneously (phase angle, amplitude, time, frequency, acoustic emission) to distinguish genuine partial discharge signals from noise. This selective triggering based on parameter combinations improves measurement precision while filtering out harmful noise signals and false positives.
2Measurement precision
If simple trigger conditions are used, then signal acquisition is fast and simple, but parallel partial discharge activities from multiple sources cannot be distinguished
Solution Approach 1:
The patent employs dimensionality change by extending the trigger condition evaluation from simple single-parameter thresholds to multi-dimensional parameter space (phase angle, amplitude, time, frequency, acoustic emission). This allows the system to distinguish multiple partial discharge sources by their unique parameter signatures while managing complexity through systematic evaluation of the expanded parameter dimensionality.
3Measurement precision
If averaging of all measurement signals is performed, then noise reduction is achieved, but signals from different sources mix and location precision deteriorates
Solution Approach 1:
The patent applies segmentation by dividing the signal processing into separate clusters based on trigger condition matching. Measurement signals are grouped into distinct clusters corresponding to different partial discharge sources, allowing noise reduction through averaging within each cluster while preventing signal mixing between sources. This maintains location precision by keeping signals from different sources separate throughout the averaging process.
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 enables precise location of partial discharge sources by averaging measurement signals from multiple sensors, reducing noise interference and accurately determining source locations even in complex transformer environments, with improved precision and reduced false positives.
Implementation Method 1
acoustic sensors for converting acoustic emission signals into electrical signals
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
Method for testing electrical components comprising the steps of: - inputting an input signal (20) having an alternating voltage to the electrical component (21), - detecting an electrical output signal that comprises partial discharge pulses due to partial discharges in the electrical component, - generating a trigger signal (7) when a partial discharge pulse is detected, and - acquiring a time-limited measurement signal sample of at least one measurement signal (8, 9, 10) in response to the trigger signal (7), characterized in that the trigger signal (7) is only generated if a physical condition defined by at least two physical properties of the input signal and/or the at least one measurement signal (8, 9, 10) are in accordance with predefined trigger condition (18, 18', 19, 19'), and that the acquired measurement signal samples are assigned to a cluster of measurement signals corresponding to the predefined trigger condition.