신경망을 이용한 유도전동기-인버터 시스템의 효율향상

Efficiency Improvement of Inverter Fed Induction Machine System Using Neural Network

  • 류준형 (아주대학교 전자공학부) ;
  • 이승철 (한국과학기술연구원 지능제어연구센터) ;
  • 최익 (한국과학기술연구원 지능제어연구센터) ;
  • 김광배 (한국과학기술연구원 지능제어연구센터) ;
  • 이광원 (아주대학교 전자공학부)
  • Ryu, Joon-Hyoung (School of Electronics Engineering, AJOU Univ.) ;
  • Lee, Seung-Chul (Intelligent System Control Research Center, KIST) ;
  • Choy, Ick (Intelligent System Control Research Center, KIST) ;
  • Kim, K.B. (Intelligent System Control Research Center, KIST) ;
  • Lee, K.W. (School of Electronics Engineering, AJOU Univ.)
  • 발행 : 1998.07.20

초록

This paper presents an optimal efficiency control for the inverter fed induction machine system using neural network. The motor speed and the load torque vary the efficiency characteristics of an induction motor. The optimal slip frequency has nonlinearity varied by the load torque as well as the motor speed. The induction motor is driven using the inverter system and the indirect vector control method which input is slip frequency. The neural network for estimating the optimal slip frequency has two input layer(the motor speed and the load torque) and one output layer(the optimal slip frequency that minimize the input power). Learning algorithm of the neural network is the back-propagation. Using the equivalent circuit including the nonlinearity of the induction motor, the loss reduction is analyzed quantitatively. Experimental results are shown noticeable power savings by proposed scheme in high speed and light load conditions.

키워드