Sensorless control of DC-DC buck converter using metaheuristic algorithm

The objective of this research work is to adjust the output voltage of a DC-DC buck converter under the influence of noise and parameter variations. For this purpose, a new Kalman filter-based fractional-order controller is proposed. The Kalman filter reduces the effects of sensor and process noise on the system output. To reduce the tuning intricacy of the fractional-order proportional-integral-derivative (FOPID) controller, circumvent the derivative term, and enhance system performance, a novel blended proportional-integral (BPI) controller is introduced. This controller combines integer-order and fractional-order proportional-integral controllers. The parameters of the proposed BPI controller are determined using four metaheuristic optimization techniques: firefly algorithm, artificial bee colony, particle swarm optimization, and Harris Hawks optimization. Among there, the potential of the firefly algorithm-based controller was superior to the other three controllers. The proposed controller is compared with the integer-order KF-based proportional-integral (PI) controller, proportional-integral-derivative (PID) controller, and KF-based fractional-order PI and PID controllers. The proposed controller presents better results regarding settling time and steady-state error. This controller also demonstrates better results under variations in input voltage and inductance of the buck converter. The results of the buck converter are compared with those from an artificial neural network (ANN)-based controller reported in previous literature. The proposed controller improves overshot by 96.42% and settling time by 40% when the inductance of the buck converter is reduced by 50% under a load change from 7.33 to 11 Ω.