Method for state diagnosis of asynchronous motor
By combining spectrum analysis and machine learning, the shortcomings of traditional MCSA in asynchronous motor fault identification are solved, and effective identification of complex mechanical faults and load changes is achieved, which improves identification accuracy and reliability and reduces computational complexity.
Patent Information
- Application Number
- CN202480017140.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-03-09
- Filing Date
- 2024-01-08
- Publication Date
- 2025-10-03
AI Technical Summary
In the existing technology of asynchronous motor fault identification, the traditional MCSA method has difficulty in effectively identifying complex mechanical faults and unknown frequencies caused by drive system oscillations, and the spectrum analysis is affected by load and slip, resulting in false alarms and insufficient identification.
A spectrum analysis-based method is adopted to calculate the spectrum variables of current and voltage through Park transform. Combined with machine learning algorithm, the spectrum amplitude and phase relationship are used as features to perform unsupervised anomaly detection and identify the fault status of asynchronous motors.
The accuracy and reliability of asynchronous motor fault identification are improved, false alarms are reduced, faults can be effectively identified under the influence of load changes and slip, and computational complexity is reduced.