Transformer DC Magnetic Bias Detection via Vibration Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting DC magnetic bias in transformers are inaccurate due to susceptibility to external influences and lack of precision, leading to potential harm and inefficiencies in transformer operation.
Innovation Solution
A method involving signal acquisition, processing, and determination steps, where vibration signals from a transformer's oil tank are analyzed through time and frequency domain decomposition to compare odd and even frequency amplitudes, facilitating accurate detection of DC magnetic bias with low costs and safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If noise and vibration methods are used to detect DC magnetic bias, then detection can be performed, but accuracy is poor due to susceptibility to external influences
Solution Approach 1:
The vibration signal is segmented into different frequency components through frequency domain decomposition. By analyzing specific frequency bands (odd harmonics at 50Hz, 150Hz and even harmonics at 100Hz), the method isolates the DC magnetic bias signature from general noise and external vibrations, thereby improving detection accuracy while reducing susceptibility to external influences.
Solution Approach 2:
Instead of directly measuring magnetic field changes, the method inverts the approach by measuring mechanical vibrations of the oil tank caused by magnetic forces, then inferring DC magnetic bias from the vibration frequency characteristics. This indirect measurement approach avoids direct exposure to magnetic interference while maintaining detection capability.
2Measurement precision
If multiple vibration sensors are arranged on the oil tank, then detection accuracy is improved, but device complexity increases
Solution Approach 1:
Multiple vibration sensors are merged into a unified signal processing framework. The vibration signals from three sensors positioned at different locations on the oil tank are combined and processed together through frequency domain decomposition, allowing the system to capture comprehensive vibration characteristics while maintaining a relatively simple overall device structure.
Solution Approach 2:
The vibration sensors serve multiple functions: they detect vibrations from DC magnetic bias, characterize the vibration patterns through frequency analysis, and provide redundant measurement data. This multi-functionality improves detection accuracy without proportionally increasing device complexity, as the same sensors perform multiple detection roles.
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 method provides a fast, safe, and highly accurate means to detect DC magnetic bias, improving transformer operation and reducing adverse effects, with the option to determine severity through amplitude ratios.
Implementation Method 1
DC magnetic bias refers to half-cycle saturation of a transformer core as a result of direct current or near-direct current entering a transformer coil from the outside and magnetizing the transformer, as well as a series of electromagnetic effects arising therefrom
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
The present invention relates to a method for detecting DC magnetic bias in a transformer, comprising the following steps: a signal acquisition step (S52), a signal processing step (S54) and a determination step (S56). The signal acquisition step consists of acquiring a vibration signal of an oil tank of the transformer. The signal processing step comprises: forming a time domain waveform of the vibration signal (S542); subjecting the time domain waveform to frequency domain decomposition, to convert the time domain waveform into a frequency domain waveform (S544); calculating an amplitude of each odd number frequency and an amplitude of each even number frequency of the frequency domain waveform (S546). The determination step is as follows: if the amplitude of an odd number frequency is greater than half of the amplitude of an even number frequency, then there is DC magnetic bias in the transformer.