| Popis: |
An adaptive model for Permanent Magnet Synchronous Motor (PMSM) is highly required in applications such as industrial, domestic, automobile, etc. The downside of PMSM is the pulsation of torque which results in instantaneous torque that pulsates periodically with the position of a rotor. As the PMSM is a synchronous machine, speed and torque are the parameters to be controlled externally. The traditional Proportional Integral (PI) controller is mainly used to control these parameters of the motor in industry because of its robustness, but the performance of PI controller tends to decline as the system opts for dynamic changes. Hence, the conventional PMSM is replaced with Artificial Neural Network (ANN) based PMSM model which will avoid the torque pulsation at the dynamic response. The main purpose of ANN is to achieve accurate prediction of the torque, current of the motor and while the speed and voltage parameters are dynamically changing. |
