Winding Temperature Perception of High-speed PMSM Based on Model and Data Dual Drive

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

The winding temperature rise of high-speed permanent magnet synchronization motor (HSPMSM) will lead to serious problems such as insulation failure, therefore it is extremely meaningful for real-time perception of winding temperature. According to the idea of model and data dual drive, this paper proposes a kind of improved recursive least square (IRLS) method based on the coordination of recursive least square (RLS) and artificial neural network (ANN). In the case of no additional signal injection, this method can effectively solve the failure problem of RLS under the condition of constant load torque. The winding temperature perception of a 20kW, 100krpm 2 pole 6 slot HSPMSM is simulated, and the simulation results demonstrate that the IRLS can correctly converge and track the temperature change of motor stator winding, and the absolute value of relative error is controlled within 1% after stability.

Original languageEnglish
Title of host publicationICEMS 2021 - 2021 24th International Conference on Electrical Machines and Systems
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages2626-2629
Number of pages4
ISBN (Electronic)9788986510218
DOIs
StatePublished - 2021
Event24th International Conference on Electrical Machines and Systems, ICEMS 2021 - Gyeongju, Korea, Republic of
Duration: 31 Oct 20213 Nov 2021

Publication series

NameICEMS 2021 - 2021 24th International Conference on Electrical Machines and Systems

Conference

Conference24th International Conference on Electrical Machines and Systems, ICEMS 2021
Country/TerritoryKorea, Republic of
CityGyeongju
Period31/10/213/11/21

Keywords

  • High-speed motor
  • improved recursive least square method
  • Winding temperature perception

Fingerprint

Dive into the research topics of 'Winding Temperature Perception of High-speed PMSM Based on Model and Data Dual Drive'. Together they form a unique fingerprint.

Cite this