Simulation of ANFIS Controller to Line Commutation Based on Current Source Converter High Voltage Direct Current System

• Published in: 2019 IEEE Conference on Energy Conversion (CENCON) pp. 84 - 89
• Main Authors: Ginarsa, I Made, Muljono, Agung Budi, Nrartha, I Made Ari
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2019
• Subjects: ANFIS; control; HVDC; percent error; simulation
• DOI: 10.1109/CENCON47160.2019.8974701

Regulation of rectifier trigger angle is very important in high voltage direct current (HVDC) transmission system to maintain effectiveness of power delivery in the HVDC operation. To cover this regulation, adaptive neuro-fuzzy inference system (ANFIS) controller is proposed in this research to replace proportional integral (PI) controller. Before applied, the ANFIS controller is trained by data training that obtained from the PI controller in various scenario. The ANFIS controller performance is evaluated by assessing its percent error of direct current compared to current reference. The result shows that at start up scenario (up-ramp 6 pu/s), the percent errors of direct current are 4.47 and 4.61% for ANFIS and conventional controllers, respectively. Also, when the current reference is reduced, the percent errors of direct current are obtained at 2.37 and 2.51% for ANFIS and conventional controllers. The ANFIS controller is able to control the HVDC with percent errors of direct current are less than the percent errors of current for PI controller in two scenarios.