Implementation of resilient directed neural network controlled virtual z source multilevel inverter fed brushless dc motor drive

Abstract

The specific features such as operations at all speeds, low rotor inertia, better dynamic characteristics and rated starting current attracts industrial applications to go for Permanent Magnet Brushless DC motors rather than Induction motor. Similarly superior efficiency, long life, smooth torque delivery and high speed operation of these motors lead the brushed DC motors in the application field such as automotive, HVAC, electronic, computer, semiconductor and medical industries too. Permanent Magnet BLDC motor implies electronic commutation through the converter circuit and switches of those converters need to be controlled with an optimum controller. Effectiveness of speed response relies on both the Converter circuit and Controller. BLDC motors are generally provided with the conventional Proportional Integral Derivative (PID) controllers fed with the traditional Voltage Source and Current Source Inverters. Two level VSI and CSI suffer from common restrictions of harmonics distortion, high DC link voltage, high dv/dt, limited output voltage, vulnerability to EMI noise and so on. Hence these traditional converters can be replaced with the combination of Z Source and Multilevel Inverter circuits. Increase in the number of steps of voltage is the solution provided by the MLI. Reliability of the converter circuit is increased by realizing ZSI which also supports in buck and boost operation. Implementation of classical controllers with simple control structure leads to high overshoot in the response. Optimal selection of converters and controllers takes the Permanent Magnet BLDC drive system with reduced peak overshoot, steady state error, rise time, settling time and good dynamic behaviour. newline