An Adaptive Neuro Fuzzy Inference System based Method for DC Fault Recognition in VSC-MTDC System

This paper presents an Adaptive Neuro Fuzzy Inference System (ANFIS) method for recognizing the fault in a Voltage Source Converter-Multiterminal HVDC (VSC-MTDC) system. A four-terminal VSC-based HVDC system is designed in MATLAB software and used for the validation of research. The proposed scheme has advanced features that overcome the limitations of a fuzzy inference system, as it does not need an expert to provide the best performance. Artificial Neural Networks (ANNs) depend only on input and output data through the training process. ANFIS is a very effective method that combines the strengths of artificial neural networks (ANN) in learning from processes and the ability of fuzzy inference systems to deal with uncertain input. In order to protect against faults, two distinct FIS models have been developed to recognize pole-to-ground and pole-to-pole faults. The results indicate a trip signal when a fault is present, hence increasing the reliability of the system. This approach provides quick outcomes without taking any feedback from the remote end of the system.

Keywords:

adaptive neuro-fuzzy inference system , DC Fault Recognition , VSC-MTDC Systems