AUTHOR=Wu Jiyang , Li Qiang , Chen Qian , Zhang Nan , Mao Chizu , Yang Litai , Wang Jinyu 

TITLE=Fault diagnosis of the HVDC system based on the CatBoost algorithm using knowledge graphs

JOURNAL=Frontiers in Energy Research

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2023.1144785

DOI=10.3389/fenrg.2023.1144785

ISSN=2296-598X

ABSTRACT=Based on the four kinds of fault data measured in a HVDC system of the Southern Power Grid, this work proposes a fault diagnosis method based on categorical boosting (CatBoost) algorithm to deal with and solve the fault problems of HVDC system. Through the verification of relevant data, it is finally proved that this method can identify the four types of faults very effectively, quickly, and accurately. Through the fault diagnosis of HVDC system, it can make great contributions to the construction and improvement of HVDC system knowledge graph (KG) in the future. In this paper, firstly, the core role and significance of fault diagnosis in KG are introduced; The characteristics and specific causes of four fault types are introduced in detail; Then sort and process the fault data, divide the fault data into training sets and test sets in proportion, and then label them; Establish the CatBoost fault diagnosis model, and substitute the training data into the model training, and then conduct the data test; Finally, the diagnostic results obtained are compared with those obtained by BP neural network algorithm. The results show that the diagnostic accuracy of CatBoost algorithm in the three test sets is always higher than that of BP neural network algorithm, and the accuracy is always higher than 94%, which fully proves that CatBoost algorithm has a very good fault diagnosis effect for HVDC system.