Novel bacterial foraging-based ANFIS for speed control of matrix converter-fed industrial BLDC motors operated under low speed and high torque

In this paper, a novel adaptive neuro-fuzzy inference system (ANFIS)-based control technique optimized by Bacterial Foraging Optimization Algorithm for speed control of matrix converter (MC)-fed brushless direct current (BLDC) motor is presented. ANFIS is considered to be one of the most promising technologies for control of electrical drives fed by MC. Optimizing the training parameters of ANFIS, to improve its performance, is still being considered by several researchers recently. Parameters of the online ANFIS controller such as learning rate (eta), forgetting factor (lambda) and steepest descent momentum constant (alpha) are optimized by using the proposed algorithm. For the purpose of comparison, proportional integral derivative controller, fuzzy logic controller, PSO-ANFIS and BAT-ANFIS are considered. Set point tracking performances of the proposed system are carried out at various operating points for an industrial BLDC motor operating at a maximum rated speed of 380 rpm and torque of 6.4 N m. Time domain specifications such as rise time, settling time, peak time, steady-state error and peak overshoot in the presence and absence of load torque disturbances are presented. Time integral performance measures such as integral square error, integral absolute error, and integral time multiplied absolute error are analyzed for various operating conditions. Speed fluctuation in the output of BLDC motor is dependent on the source current harmonics of the inverter/converter. To illustrate this, total harmonic distortion (THD) analysis is carried out for the existing PWM inverter and the proposed MC, and it is proved that MC results in reduced THD, as compared to PWM inverter. Simulation results confirm that the proposed controller outperforms the other existing control techniques under various set speed and torque conditions. Statistical analysis is effectively carried out to prove the effectiveness of the proposed controller. Experimental analysis is performed to validate the performance of the proposed control scheme.