SENSORLESS DIRECT TORQUE CONTROLLED DRIVE OF BRUSHLESS DC MOTOR BASED ON FUZZY LOGIC

INTRODUCTION
Permanent magnet brushless DC motor (BLDC) with trapezoidal back-EMF have been widely used in many field of variable-speed drives for their higher power/weight and higher efficiency. However, in practice, torque ripple may exist for the motor itself and feeding system. Particularly, torque ripple during the commutation period is one of the main drawbacks which deteriorate the performance of BLDC drives. To reducing torque ripple effectively, various torque control methods. In, a novel torque controller attenuating the undesired torque pulsation for BLDC with non-ideal trapezoidal back- EMF is presented, in which the torque is estimated from the product of the instantaneous back-EMF and current. However, the winding resistance was neglected and the back-EMF shape functions according to the rotor position are tested by offline, and set up at the look up table. 

For achieving instantaneous torque control and reducing the torque ripple, direct torque control (DTC) has been successfully extended to a three-phase BLDC drive operating in the 120 elec. Degrees conduction mode. DTC scheme was originally developed for induction machines drives which was first proposed by Takahashi and Depenbrock in the mid 1980s. Control of torque is exercised through control of the amplitude and angular position of the stator flux vector relative to the rotor flux vector. It is claimed that the electromagnetic torque and the amplitude of stator flux linkage can be controlled simultaneously. However, the control effect of the stator flux linkage is not good from the simulated and experimental result.  In this paper, a sensorless fuzzy direct torque controlled BLDC drives is proposed to reducing torque ripple in twophase conduction mode. The proposed scheme differs from the direct torque controlled BLDC drives in that the amplitude of stator flux linkage is not controlled since it is automatically determined by every 60 elec. degrees commutation. In the system, the torque error and flux linkage angle of BLDC were all properly fuzzified into several subsets to accurately select the voltage space vector in order to smooth the torque and quicken the torque response. The torque and rotor speed are derived from the back-EMFs, which are estimated from a designed state observer. Its effectiveness is validated by simulations. 

IMPLEMENTATION OF FUZZY BLDC-DTC SYSTEM
The block diagram of a sensorless BLDC drive with fuzzy DTC may be as shown in Fig.

VIDEO DEMO