PMSM Torque Ripple Reduction in Electric Vehicle using Neural Network

For Electric vehicle(EV) application, Permanent Magnet Synchronous Motor (PMSM) is widely used due to high power density and high efficiency. Field Oriented Control (FOC) with feedforward compensation is used predominantly for motor control to give better dynamic performance. With the changing motor...

Full description

Saved in:
Bibliographic Details
Published in2021 IEEE International Power and Renewable Energy Conference (IPRECON) pp. 1 - 6
Main Authors Kuvalekar, Suryakant A., Mohanrajan, S.R.
Format Conference Proceeding
LanguageEnglish
Published IEEE 24.09.2021
Subjects
Artificial neural networks
Electric Vehicle
FOC
Neural Network
Permanent magnet motors
PI control
PMSM
Simulation
Synchronous motors
Torque
Torque control
Online AccessGet full text

Cover

More Information
Summary:For Electric vehicle(EV) application, Permanent Magnet Synchronous Motor (PMSM) is widely used due to high power density and high efficiency. Field Oriented Control (FOC) with feedforward compensation is used predominantly for motor control to give better dynamic performance. With the changing motor parameters due to ageing effect or variation in motor temperature causes torque ripple and EV vibrations. This paper presents the implementation of Neural Network (NN) for PMSM Control to reduce the torque ripples. NN with current feedback works as a feedforward network. Current control PI regulators and feedforward compensation is replaced with NN model. It improves the decoupling accuracy in between d-axis and q-axis currents and also reduces the torque ripples even if motor parameters varied slightly. Vehicle dynamics is taken into consideration during simulation. Matlab/Simulink tool is used for simulation and verified the Motor torque performance with FOC and NN.
DOI:10.1109/IPRECON52453.2021.9641002