Stationary Signal Segmentation for Variable-Speed Drive Diagnostics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for condition monitoring of electromechanical systems, such as Motor Current Signature Analysis (MCSA), are less effective when electrical rotating machines are supplied by variable-speed-drives, as they struggle with non-stationary current signals and frequency variations, leading to difficulties in distinguishing peaks of interest from noise and harmonics.
Innovation Solution
A method that involves measuring an analog waveform signal, converting it to a discrete processed signal, subdividing and resampling it to create stationary periods, and then transforming it into the frequency domain for analysis, allowing for the extraction of interest frequencies and amplitudes without requiring prior knowledge of the frequency content, thus enabling effective diagnostics regardless of supply frequency variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Motor Current Signature_analysis is applied to variable-speed-drive supplied machines, then diagnostics can be performed, but the non-stationary nature of current signals causes peaks of interest to cease to occur at single distinct frequencies and become difficult to distinguish from noise
Solution Approach 1:
The patent divides the non-stationary current signal into multiple stationary segments by identifying zero-crossings and creating discrete periods. Each segment is then processed independently through Fourier transform, allowing frequency analysis to be performed on otherwise non-stationary data. This segmentation transforms the problematic continuous non-stationary signal into manageable stationary pieces that can be analyzed using standard spectral methods.
Solution Approach 2:
The patent dynamically adjusts the analysis approach by continuously monitoring the signal characteristics and adapting the segmentation and processing parameters in real-time. The system identifies zero-crossings dynamically and creates segments based on the instantaneous signal state, allowing the analysis to adapt to changing operating conditions without requiring prior knowledge of the signal content.
2Adaptability or versatility
If adjustable frequency clock circuitry is used to sample current signals, then sampling rate can be adjusted to match supply frequency, but the circuitry becomes much more complex and introduces lag and noise susceptibility
Solution Approach 1:
The patent replaces the mechanical/electronic adjustable frequency clock system with a software-based signal processing approach. Instead of using complex hardware circuitry to dynamically adjust sampling rates, the system uses digital signal processing techniques including zero-crossing detection, segment creation, and discrete Fourier transforms to achieve frequency analysis. This substitution eliminates the hardware complexity and associated lag and noise problems while maintaining adaptability to varying supply frequencies.
Data Source
AI summary
The present invention is concerned with a method of determining stationary signals for the diagnostics of an electromechanical systems in which electrical rotating machinery is used and in which at least one electrical or mechanical signal is measured during an operation of the electromechanical system. The method is used especially for condition monitoring of electric motors and generators. The method consists ofmeasuring an analog waveform signal (S) of the electromechanical systemand then manipulating that signal in various ways to obtain a frequencies spectrum, from which a vector of interest frequencies and corresponding vector of amplitudes are extracted to diagnose the electromechanical system.