67088

Автор(ы): 

Автор(ов): 

4

Параметры публикации

Тип публикации: 

Доклад

Название: 

On Speed Controller Neural Tuner Application to Compensate PMSM Mechanics Inertia Moment Drift

Электронная публикация: 

Да

ISBN/ISSN: 

03029743

DOI: 

10.1007/978-3-319-92537-0_83

Наименование конференции: 

  • 15th International Symposium on Neural Networks (ISNN 2018, Minsk)

Наименование источника: 

  • Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Обозначение и номер тома: 

Vol. 10878

Город: 

  • Минск

Издательство: 

  • Springer

Год издания: 

2018

Страницы: 

727-735
Аннотация
A problem of adaptive control methods application to a speed vector control system of a Permanent Magnet Synchronous Motor (PMSM) with time varying mechanical parameters is considered. Such methods analysis is made to select the most appropriate one for the problem under consideration. As a result, a neural tuner is chosen. A synchronous motor mathematical model is shown, and the vector based control system of the motor is described. The neural tuner structure and its operation principle are presented. It is applied to adjust the speed controller of the PMSM. Experiments are conducted using the mathematical model of a PMSM Siemens 1FK7103, in which the mechanical inertia moment value is changed gradually during the modeling process. The tuner application as an augmentation to the speed controller allows to keep the required transient quality during the experiment despite the drive nonstationarity in contrast to a P-controller with the constant parameter value. This results in avoidance of, firstly, the transient time increase when the inertia moment is higher than its nominal value and, secondly, the speed overshoot in vice versa case.

Библиографическая ссылка: 

Еременко Ю.И., Глущенко А.И., Петров В.А., Молодых А.В. On Speed Controller Neural Tuner Application to Compensate PMSM Mechanics Inertia Moment Drift / Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Минск: Springer, 2018. Vol. 10878. С. 727-735.