A Single Neuron Speed Controller for BLDC Motor Driven by Pseudo Sinusoid Commutation

The paper presents a novel method of speed control with Single neuron PI controller with a simple and novel commutation scheme, "Pseudo Sinusoid Commutation" for a BLDC motor with improved performance. BLDC motors are one of the electrical drives that are rapidly gaining popularity, due to their high efficiency, good dynamic response and low maintenance. The proposed scheme of commutation is theoretically a compromise between six-step trapezoidal commutation and the sinusoidal commutation. The proposed algorithm has considerably lesser torque ripple in comparison to the existing trapezoidal commutation. In addition to the proposed commutation scheme addresses the inherent disadvantage of higher switching losses associated with sinusoidal commutation. The proposed commutation scheme utilizes the six non-zero switching vectors of the 2-level inverter to synthesize the reference vector at any instant. The proposed commutation scheme can be implemented in analog domain without complexity. A single neuron PI controller is used to control the speed, by adjusting the weight. The simulations for the schemes presented in the paper are carried out in MATLAB/SIMULINK, according to simulation, the proposed strategy shows good self adapted track ability, also the structure of the drive is simplified.