Experimental evaluation of dynamic performance of three‐phase AC–DC PWM rectifier with PD‐type‐2 fuzzy neural network controller
Hakan Açıkgöz, Resul Çöteli̇, Beşir Dandıl, Fikret Ata · IET Power Electronics · 2018
Diode and thyristor‐based rectifier circuits have been widely used in the industry. Due to non‐linear structures of these circuits, they draw non‐sinusoidal current from AC network as well as cause a low power factor in the AC side. The DC‐link voltage of rectifier is affected by the changes in AC network or by the load variations on the DC side. Pulse‐width modulated (PWM) rectifiers can eliminate the mentioned power quality problems if they control properly. This study proposes a controller with an adaptive and robust structure based on proportional + derivative type‐2 fuzzy neural network (PD‐T2FNN) for DC‐link voltage control of PWM rectifier. Dynamic performance of PWM rectifier using the proposed controller is evaluated via dSPACE based experimental setup under different operation conditions: set‐point change, step load change in the DC side of the rectifier, set‐point change under load and capacitive operation mode. The experimental results are given for traditional PD and proportional + integral and T2FNN controllers to validity performance of the proposed controller. Performances of controllers are evaluated regarding settling time, overshoot, steady‐state error and total harmonic distortion. PWM rectifier with PD‐T2FNN DC‐link voltage controller has superior performance for all operating conditions according to performance criteria when compared with other controllers.