hybrid ai for enhanced voltage control in three-phase boost rectifiers

Objective: three-phase boost rectifier is a voltage-source converter that converts three-phase ac input voltage to a higher dc voltage. in this paper, an artificial intelligent-based system, with learning and adapting ability, is designed for using in the two voltage-based control methods of rectifiers, with the names of voltage oriented control (voc) and direct power control (dpc). for implementation of this intelligent controller a hybrid structure of the fuzzy logic (fl) and neural networks (nn) that named as adaptive network-based fuzzy inference system (anfis) is used. among the common network training algorithms, the error back propagation algorithm is known as the most common solution by providing an efficient computational method, so in this article, the above method is used to design the controller. this neuro-fuzzy-based control model is applicable in both voc and dpc methods and increases the correctness of the output current and dc voltage with low ripple, short settling time and also dynamic operation. the implementation of the proposed controller is simple and requires only 49 fuzzy rules. compared to other controllers whose structure is neural network and fuzzy, it has fewer layers and its accuracy is higher.