MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202441099855 A) filed by Aman Chouhan; Satya Prakash; and Dr. S. Nirmal Raj on December 17, 2024, for Machine Learning Techniques For Classification Of Tolerance In Pmsm Motor.
Inventors include Aman Chouhan; Satya Prakash; and Dr. S. Nirmal Raj.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: Permanent Magnet Synchronous Motors (PMSMs) are critical components in various industrial applications due to their high efficiency, power density, and reliability. However, the performance and longevity of PMSMs can be affected by manufacturing and operational tolerances. Accurate classification of these tolerances is essential for predictive maintenance, quality control, and optimizing motor performance. Machine learning (ML) techniques have emerged as powerful tools for classifying tolerances in PMSM motors, enabling more precise diagnostics and monitoring. This paper explores various ML approaches, including supervised and unsupervised learning algorithms, to classify tolerances based on data obtained from PMSM operation. Key methods such as Support Vector Machines (SVM), Random Forests, and Neural Networks are analyzed for their effectiveness in handling the complex and nonlinear characteristics of PMSM tolerance data. The study also evaluates the impact of feature selection, data preprocessing, and model optimization techniques on classification accuracy. The results demonstrate that ML-based approaches can significantly enhance the identification and classification of tolerances, leading to improved motor performance and reduced downtime.
Disclaimer: Curated by HT Syndication.
