A SMOTE-Tomek-Based Parameter Identification and Behavior Estimation Method for IPMSM in Aerial Applications

In aerial applications, Interior Permanent Magnet Synchronous Motors (IPMSMs) are widely utilized due to their high power density and efficiency. However, accurate parameter identification and behavior estimation of IPMSMs are challenging tasks, especially under varying operating conditions and uncertainties. This paper proposes a novel SMOTE-Tomek-based method for parameter identification and behavior estimation of IPMSMs in aerial applications. The Synthetic Minority Over-sampling Technique (SMOTE) is employed to address the class imbalance issue in the dataset, while Tomek links are utilized to remove noisy samples that can negatively affect the identification and estimation performance. The proposed method first collects a large open-source dataset of both SPMSM (Surface-mounted PMSM) and IPMSM operating conditions, including various speeds, loads, and temperatures. After preprocessing the dataset, a SMOTE-Tomek approach is trained to identify the IPMSM parameters and a convex hull is dedicated to estimating its behavior accurately. The trained model is capable of identifying IPMSM parameters such as flux linkage, resistance, and inductance. Extensive experiments and simulations are conducted to evaluate the performance of the proposed SMOTE-Tomek-based method in an IPMSM case study. The proposed method has the potential to enhance the performance and reliability of IPMSM-driven aerial applications, including unmanned aerial vehicles, drones, and electric aircraft.