Inductance Parameter Online Identification of PMASRM Based on Current Harmonics Extraction with Ultra-local Model
ID：40
Submission ID：44 View Protection：ATTENDEE
Updated Time：2025-10-11 22:29:54 Hits：67
Poster Presentation
Start Time：2025-11-09 09:02 （Asia/Shanghai）
Duration：1min
Session：[P] Poster presentation » [P7] 7.Electric Machine Design and Control
No files
Abstract
The traditional deadbeat predictive current control (DPCC) is renowned for its excellent performance, quick reaction, simplicity, and ease of use. DPCC, on the other hand, is extremely sensitive to changes in machine parameters, and when model parameters diverge from the real values, performance suffers greatly. This issue is particularly pronounced in permanent magnet assisted synchronous reluctance motors (PMASRMs), where the inductance decreases as the current increases. As a result, the DPCC method, which typically measures motor inductance under no-load conditions, will encounter considerable performance degradation with increasing current levels. In order to address this problem, this work proposes a DPCC approach based on the ultra-local model, which only uses system input and output and does not require motor parameter information, hence improving control system robustness. Additionally, to enable real-time motor inductance identification, an improved high-frequency current injection method is put forward, which introduces a current harmonic amplitude extraction link and replaces the low-pass filter with average value filter. By using this method, the accuracy and convergence rate of inductance identification can be greatly enhanced. Finally, the efficacy of the suggested control strategy for PMASRM is confirmed by comparison of simulations.
Keywords
Extended state observer (ESO), high-frequency current injection, online full parameter identification, permanent magnet assisted synchronous reluctance motor (PMASRM), ultra-local model
Speaker
Comment submit