Sensorless speed and position control of permanent magnet BLDC motor using particle swarm optimization and ANFIS

Abstract

This paper describes the operation of a Permanent Magnet Brushless Direct Current (PMBLDC) motor without a position sensor. In this case, the sensorless operation is enhanced by an effective hybrid technique that detects the back electromotive force (Back EMF) of the zero crossing point (ZCP) from the terminal voltages. The Adaptive Neuro Fuzzy Interference System (ANFIS) controller, which is based on Particle Swarm Optimization (PSO) and uses PSO to train its operation, is combined in the proposed hybrid analysis. The PMBLDC motor's ANFIS controller receives the line voltages as input, and it uses this information to estimate the sample signals that are then sent to the ZCP detection circuit. Appropriate commutation control of the inverter is generated by the ZCP detecting circuit. By varying the ANFIS consequent parameters, the PSO algorithm iterates until the error between the sample output and the real training data reaches a low value. The MATLAB/Simulink platform is utilized to implement the suggested sensorless controller action. To verify the controller's performance, a comparison with the other soft computing methods is also carried out.