Linear Hall Eccentricity Diagnosis for Permanent Magnet Motors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting rotor eccentricity in permanent magnet synchronous motors face challenges such as difficulty in data collection, high hardware costs, and inability to accurately distinguish between static and dynamic eccentricity, especially for motors with different topologies.
Innovation Solution
A linear Hall-based eccentricity diagnosis method using three Hall elements mounted in stator slots with a digital signal processor, employing a complex factor filter and phase-locked loops to process output signals, allowing for real-time separation of static and dynamic eccentricity detection in permanent magnet motors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional windings are used for eccentricity detection, then the detection capability is improved, but the voltage amplitude varies greatly with rotational speed making data collection difficult and increasing hardware cost
Solution Approach 1:
The patent replaces the mechanical/electrical additional winding system with a magnetic field-based detection system using Hall sensors. This substitution eliminates the need for speed-proportional voltage generation, allowing eccentricity detection without the amplitude variation problem inherent in winding-based methods. The Hall sensors directly measure magnetic field distribution, providing speed-independent detection data.
Solution Approach 2:
The patent introduces Hall sensors as intermediary devices that indirectly measure eccentricity through magnetic field detection rather than directly measuring electrical parameters. This intermediary approach allows detection without direct electrical connection to rotating parts, simplifying the system while maintaining detection accuracy across varying speeds.
2Adaptability or versatility
If 2N Hall sensors are mounted in two stator slots radially symmetrical to the center axis, then hub motor eccentricity and demagnetization fault can be decoupled diagnosed, but accurate detection in specific static eccentric state cannot be achieved
Solution Approach 1:
The patent divides the detection task into separate functional components: three Hall sensors arranged at specific angular intervals (120 degrees) independently measure magnetic field components, and the detection algorithm processes these segmented measurements to separately identify static and dynamic eccentricity. This segmentation allows accurate detection in specific static eccentric states while maintaining decoupling capability.
Solution Approach 2:
The patent employs an asymmetric sensor arrangement with three Hall sensors positioned at specific angular intervals rather than symmetric placement. This asymmetric configuration creates distinct magnetic field measurement patterns that enable the detection algorithm to differentiate between static and dynamic eccentricity components, achieving accurate detection in previously problematic static eccentric states.
3Measurement precision
If neural network model is used for static eccentricity detection, then detection can be performed, but computational burden increases and robustness decreases due to errors between model and actual motor parameters
Solution Approach 1:
The patent replaces the complex, computationally intensive neural network model with a simpler, analytically-based detection algorithm. This simpler approach uses direct mathematical relationships between Hall sensor measurements and eccentricity parameters, eliminating the need for heavy computational resources while maintaining detection accuracy and improving robustness against parameter variations.
Solution Approach 2:
The patent changes the detection approach from data-driven neural network parameter estimation to physics-based analytical parameter calculation. By using established magnetic field equations and direct mathematical relationships, the system achieves accurate eccentricity detection without the computational burden and parameter sensitivity issues inherent in neural network models.
4Adaptability or versatility
If existing detection methods are used, then some form of eccentricity detection is possible, but the ability to realize simultaneous detection of static and dynamic eccentricity for motors of different topologies is lacking
Solution Approach 1:
The patent creates a universal detection system using three Hall sensors arranged at 120-degree intervals that can simultaneously detect both static and dynamic eccentricity across different motor topologies. The detection algorithm is designed to extract both eccentricity components from the same sensor measurements, providing multi-functional capability that works for various motor configurations without requiring topology-specific adjustments.
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, low-cost, and non-invasive eccentricity detection for motors with various topologies by rapidly processing output signals from Hall sensors, effectively separating static and dynamic eccentricity indicators.
Implementation Method 1
three linear Hall elements mounted in stator slots at the same space interval... radial flux density... output voltage signals
Data Source
AI summary
A linear Hall-based eccentricity diagnosis method and detection system for a permanent magnet motor. First, three linear Hall elements are mounted in stator slots at the same space interval, respectively; second, analog signals output by the three-phase linear Hall are converted into digital signals by means of a digital signal processor, and the digital signals are converted into a quadrature signal by means of linear combination; then, a negative sequence signal and a sideband signal are extracted from the quadrature signal by means of a complex factor filter; then the amplitude of the negative sequence signal and the amplitude of the sideband signal are extracted by means of synchronous reference frame phase-locked loops as a static eccentricity indicator and a dynamic eccentricity indicator; finally, percentages representing the degrees of eccentricity are calculated from the static eccentricity indicator and the dynamic eccentricity indicator in the digital signal processor.