Modeling dynamic characteristics of synchronous motor with radial based artificial neural networks


Creative Commons License

Gani A., Doğmuş O., Açıkgöz H., Keçecioğlu Ö. F., Yıldız C., Şekkeli M.

1st International Mediterranean Science and Engineering Congress (IMSEC 2016), Adana, Türkiye, 26 - 28 Ekim 2016, ss.1-11

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Basıldığı Şehir: Adana
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.1-11
  • Kayseri Üniversitesi Adresli: Hayır

Özet

System modelling can be defined as system models that created thanks to the data obtained from applications or mathematical expressions. Modelling helps determine and design dynamic behaviors of a system. Most of industrial applications are not linear. In recent years, artificial neural networks have been widely used in the modelling of complex and non-linear systems. Radial basis artificial neural network is an artificial neural network used in the modelling of non-linear systems. They have increased their popularity in recent years due to their success in curve fitting and classifying non-linear problems. kaTherefore, studies on software and hardware applications related to radial basis artificial neural networks have gained momentum. Synchronous motor is an alternative current motor in which rotor rotational speed is equal to the rotational speed of the stator rotating field and the rotation speed does not vary in loading. These motors are used in industrial applications requiring steady speed. When excitation current of the synchronous motor changes, it is fed by ohmic, inductive and capacitive current. In a synchronous motor operating at a constant load and voltage, the characteristic which yields the relationship between excitation current and stator current is called V current. The change in excitation current is inversely proportional to the change in power coefficient under the same conditions. This study focuses on modelling excitation current-stator current characteristic (V current) and excitation current-power coefficient characteristic based on the data obtained from simulation studies on changes in the excitation current of synchronous motor using radial basis artificial neural networks, which is a special multi-layered and feed-forward form of artificial neural networks.