Partial Discharge Signal Correlation in Three-Phase Machines

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods fail to reliably measure and diagnose partial electrical discharges in three-phase electric machines powered with square wave voltage, particularly in new-generation low-voltage converters, due to high-frequency interference and the inability to correlate discharge signals with the corresponding power supply voltages, leading to inaccurate diagnostic assessments.

Innovation Solution

A method that transforms sets of three voltage values into pairs, derives phase angle values related to converter logic states, and attributes these phase angles to discharge signals to assess and separate them based on corresponding phase values, enabling correlation with the power supply voltages and distinguishing between noise and actual discharges.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a single sensor is used to measure partial discharge signals in all three phases, then device complexity is reduced, but the ability to reliably distinguish discharge signals from interference is worsened

Engineering Contradiction:
Improvesensor configurationVSAvoiddischarge signal detection accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent segments the measurement data by transforming three-phase voltage values into phase angle values, allowing the single sensor's discharge signals to be categorized and associated with specific phases. This segmentation enables reliable distinction between actual discharges and interference despite using a single sensor.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces phase angle values as an intermediary parameter that mediates between the single sensor's raw discharge signals and the three-phase power supply voltages. By deriving phase angles from voltage measurements and attributing them to discharge signals, the system enables accurate signal identification without requiring multiple sensors.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If phase angle attribution is implemented to correlate discharge signals with power supply voltages, then measurement precision is improved, but device complexity and processing time are worsened

Engineering Contradiction:
Improvedischarge signal correlation accuracyVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes the parameter representation by transforming three-phase voltage values into phase angle values. This parameter transformation simplifies the correlation process between discharge signals and power supply voltages, improving measurement precision while keeping processing manageable through mathematical transformation.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent performs preliminary transformation of voltage values into phase angles before correlating them with discharge signals. This preliminary action organizes the voltage data in a form that facilitates efficient and accurate correlation with measured discharge signals, reducing overall processing complexity.

Inventive Principle:
Principle #10Preliminary action

3Productivity

If statistical processing is performed on heterogeneous discharge data from all phases, then productivity is improved, but measurement precision is worsened due to mixing discharge signals with interference

Engineering Contradiction:
Improvedata processing efficiencyVSAvoiddiagnostic accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent segments discharge signals into homogeneous classes based on attributed phase angles before statistical processing. This segmentation separates actual discharge signals from interference and organizes data by phase, ensuring that statistical analysis is performed on comparable data, thereby maintaining both productivity and measurement precision.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent groups discharge signals into homogeneous classes based on their attributed phase angles. This homogenization ensures that statistical processing is applied to signals from the same phase and similar conditions, improving diagnostic accuracy while maintaining processing efficiency through systematic organization.

Inventive Principle:
Principle #33Homogeneity

Data Source

PatentEP2109777B1A method for processing data pertaining to an activity of partial electrical discharges
Publication Date: 2010.11.24 TECHIMP TECHNOLOGIES SA
  • EP2109777B1 patent drawingFigure 1~2
  • EP2109777B1 patent drawingFigure 3~5
  • EP2109777B1 patent drawingFigure 6~7

AI summary

A method for measuring and processing data pertaining to an activity of partial electrical discharges taking place in a three-phase electric machine (1) powered with square wave voltage, said data comprising a plurality of discharge signals and a plurality of sets of three values (eu, Qv, ew) of power supply voltage of the machine measured in concurrence with a corresponding discharge signal, comprises the following steps: - transforming each set (eu, Qv, e\?) of three voltage values measured into a corresponding pair (ed, eq) of voltage values; - deriving a phase angle value from each of said pairs (ßd, eq) of voltage values; - attributing to each discharge signal the corresponding phase value, to assess the measured discharge signals as a function of corresponding phase values. Prior art techniques for processing data pertaining to activity of partial electrical discharges taking place in three phase electric machines powered with square wave voltage provide for assessing the measured discharged signals as a time sequence of events, with no possibility to place them effectively in relation to the power supply voltages, this entailing that any subsequent statistical processing of the measured data will have little significance.