OLTC Vibration Signature Analysis for Noise-Robust Deviation Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for monitoring electrical switching components immersed in a liquid, such as on-load tap changers, face challenges in accurately assessing deviations in switching times due to environmental noise and interference, which can lead to malfunction or wear, making condition-based maintenance difficult.

Innovation Solution

A computer-implemented method using a vibration signature model and dual smoothing levels to analyze vibration signals from electrical switching components, comparing smoothed curves to identify deviations and provide a deviation status, enhancing accuracy by filtering out noise and ensuring correct peak matching.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If vibration analysis is used to monitor electrical switching components, then condition-based maintenance capability is improved, but measurement precision deteriorates due to environmental noise and interference

Engineering Contradiction:
Improvecondition-based maintenance capabilityVSAvoidvibration signal accuracy
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The vibration signal processing is segmented into multiple stages: raw signal acquisition, noise filtering, peak detection, and deviation analysis. This segmentation allows each stage to focus on specific aspects of signal processing, improving overall measurement precision while maintaining reliability

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Signal processing algorithms act as intermediaries between the raw vibration signals and the final analysis results. These intermediaries filter out environmental noise and interference, allowing accurate detection of switching component conditions without being affected by external factors

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If dual smoothing levels are applied to vibration signals, then measurement precision is improved by filtering noise, but device complexity increases

Engineering Contradiction:
Improvedeviation detection accuracyVSAvoidsignal processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

Two different smoothing levels are applied to the vibration signal - a first smoothing level for overall trend identification and a second smoothing level for detailed peak analysis. This partial application of different processing intensities improves measurement precision without requiring complete reprocessing of the entire signal at maximum complexity

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The first smoothing level is applied preliminarily to the raw vibration signal to remove high-frequency noise before the second smoothing level is applied for detailed peak detection. This preliminary action simplifies subsequent processing steps and reduces overall computational complexity

Inventive Principle:
Principle #10Preliminary action

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

The method improves the accuracy of monitoring electrical switching components by reducing noise interference, enabling precise identification of deviations in switching times and sequences, thus facilitating effective condition-based maintenance.

Implementation Method 1

a vibration sensor configured to measure vibrations of the container

Methodology Applied
Scientific EffectVibration: Vibration

Data Source

PatentEP4498047B1A computer-implemented method for monitoring an electrical switching component immersed in a liquid
Publication Date: 2025.11.05 HITACHI ENERGY LTD
  • EP4498047B1 patent drawingFigure 1
  • EP4498047B1 patent drawingFigure 2
  • EP4498047B1 patent drawingFigure 3

AI summary

A computer-implemented method (M1) for monitoring an electrical switching component (1) immersed in a liquid in a container (2) of a transformer or of an on-load tap changer (OLTC) (3), said method (M) comprising, by a processing circuitry (4) of a computer system (5) obtaining (M1) a vibration signature model, obtaining (M2) a vibration signal recording from a vibration sensor configured to measure vibrations of the container, determining (M3) a first smoothed curve (C1) based on the vibration signal recording using a first smoothing level, determining (M4) a second smoothed curve (C2) based on the vibration signal recording using a second smoothing level lower than the first smoothing level, (M5) selecting a portion in time of the second smoothed curve based on an extent in time of a peak of the first smoothed curve (C1), comparing (M6) one or more peaks of the selected portion the second smoothed curve to the one or more peaks of a corresponding portion of the vibration signature to determine one or more deviations based on the comparison, determining (M7) a deviation status based on the deviations, and providing (M8) a deviation status signal based on the deviation status.