Optimal parameters estimation of a BLDC motor by Kohonen's Self Organizing Map Method

Brushless DC motors are the widely used motors for they possess many advantages when compared with induction motors such as higher efficiencies, High torque to inertia ratios, Greater speed capabilities, Lower audible noise, Better thermal efficiencies, Lower EMI characteristics, electronically commutated etc., In the design of such advantageous motors it becomes necessary for the estimation of the performance characteristics parameters such as back EMF, stator current, rotor speed, Torque etc., Many ideas have been proposed for the estimation of these characteristic parameters. This paper proposes an unsupervised learning method i.e., Kohonen's Self Organizing Feature Map method of estimation of BLDCM drive parameters. Since the method makes use of `winner takes it all' of neurons, the values obtained by this, will be the optimal values. Simulation of the drive is first performed under ideal conditions and the values of the above mentioned parameters are obtained. Matlab coding is then written for KSOFM which is run and various maps of KSOFM are obtained. The values obtained using these two methods are compared and is found to match with each other. Because of the idea of "Winner takes it all" and the comparison with the ideal simulation, it can be concluded that the values obtained are optimal. As mentioned Matlab/Simulink is used for the simulation and the results obtained are shown with the inferences.