Motor Current FFT Diagnosis with Inverter Noise Peak Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In systems with inverters for electric motor drive control, the spectrum peaks due to the inverter, motive power transmission mechanism, and electric motor abnormalities often overlap, leading to inaccurate abnormality diagnosis.
Innovation Solution
An abnormality diagnosis device and method that uses FFT analysis to distinguish between spectrum peaks caused by inverter noise and those from the motive power transmission mechanism, employing a monitoring diagnosis unit with peak analysis and frequency determination units to separate and identify genuine motor abnormalities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If FFT analysis is performed on current to diagnose abnormality, then abnormality detection capability is improved, but spectrum peaks from inverter noise overlap with motor abnormality peaks causing false diagnosis
Solution Approach 1:
The patent segments the spectrum analysis process into distinct stages: initial learning phase where only inverter noise peaks are identified and stored, followed by a diagnosis phase where these stored peaks are subtracted from new measurements. This segmentation separates the identification of noise characteristics from the actual abnormality detection, preventing overlap between inverter noise and motor abnormality peaks.
Solution Approach 2:
The patent performs preliminary action by conducting an initial learning phase before actual diagnosis begins. During this phase, the system learns and stores the characteristic frequencies of inverter noise peaks without performing abnormality determination. This preliminary identification of noise patterns enables subsequent subtraction of these peaks from diagnostic measurements, eliminating false positives.
2Ease of operation
If inverter is used for drive control, then motor control performance is improved, but inverter noise peaks overlap with motor and transmission mechanism spectrum peaks
Solution Approach 1:
The patent extracts inverter noise peaks from the overall spectrum by identifying them during an initial learning phase and storing their characteristic frequencies. During subsequent diagnosis, these extracted noise peaks are subtracted from the measured spectrum, effectively removing the inverter noise component and leaving only the motor and transmission mechanism peaks for accurate analysis.
3Device complexity
If conventional spectrum analysis is used without inverter noise compensation, then analysis simplicity is maintained, but erroneous determination occurs due to overlapping peaks
Solution Approach 1:
The patent performs preliminary action by conducting an initial learning phase before actual diagnosis begins. During this phase, the system learns and stores the characteristic frequencies of inverter noise peaks without performing abnormality determination. This preliminary identification of noise patterns enables subsequent subtraction of these peaks from diagnostic measurements, eliminating false positives.
Solution Approach 2:
The patent implements feedback by using the initially learned inverter noise peak frequencies to adjust and refine subsequent diagnostic measurements. The stored noise characteristics provide a reference that feeds into the diagnosis process, enabling the system to compensate for inverter noise effects and improve diagnosis accuracy through iterative refinement.
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
Enables accurate abnormality diagnosis by extracting and isolating inverter noise peaks, preventing erroneous determinations and ensuring precise identification of motor and transmission mechanism issues.
Implementation Method 1
a current detection circuit for detecting current of the electric motor
Implementation Method 2
spectrum peaks extracted by performing FFT analysis of current detected by the current detection circuit
Data Source
AI summary
An abnormality diagnosis device performs determination for at least one of abnormality of an electric motor driven by an inverter driven at a predetermined operation frequency and abnormality of a motive power transmission mechanism which transmits motive power from the electric motor to a load, wherein spectrum peaks extracted through FFT analysis of detected current of the electric motor are analyzed, and frequencies of spectrum peaks due to noise of the inverter are acquired in advance using the operation frequency and the frequencies of sideband waves with respect to the operation frequency. In abnormality diagnosis, abnormality determination is performed after spectrum peaks due to noise of the inverter are extracted from spectrum peaks extracted through FFT analysis of detected current of the electric motor.