Studi diagnosis kegagalan Transformator Daya berbasis Dissolved Gas Analysis menggunakan Algoritma K-Nearest Neighbors

Hidayat, Rafi Maulana (2024) Studi diagnosis kegagalan Transformator Daya berbasis Dissolved Gas Analysis menggunakan Algoritma K-Nearest Neighbors. Sarjana thesis, UIN Sunan Gunung Djati Bandung.

Preview	Text (COVER) 1_cover.pdf Download (213kB) | Preview
Preview	Text (ABSTRAK) 2_abstrak.pdf Download (247kB) | Preview
Preview	Text (DAFTAR ISI) 3_daftarisi.pdf Download (219kB) | Preview
Preview	Text (BAB I) 4_bab1.pdf Download (317kB) | Preview
	Text (BAB II) 5_bab2.pdf Restricted to Registered users only Download (826kB) | Request a copy
	Text (BAB III) 6_bab3.pdf Restricted to Registered users only Download (368kB) | Request a copy
	Text (BAB IV) 7_bab4.pdf Restricted to Registered users only Download (420kB) | Request a copy
	Text (BAB V) 8_bab5.pdf Restricted to Registered users only Download (473kB) | Request a copy
	Text (BAB VI) 9_bab6.pdf Restricted to Registered users only Download (195kB) | Request a copy
	Text (DAFTAR PUSTAKA) 10_daftarpustaka.pdf Restricted to Registered users only Download (214kB) | Request a copy

Abstract

Dissolved Gas Analysis (DGA) is a method to identify the type of failure in a transformer by assessing the amount of gas contained in the transformer's insulating oil. DGA has several methods of analyzing and identifying failure types based on the type of gas dissolved. However, with large amounts of data this method becomes difficult and requires expertise in graphical failure detection. This research aims to improve the diagnostic accuracy of transformer failures by implementing the K-Nearest Neighbours (KNN) algorithm on each conventional DGA method namely Roger Ratio, Duval Triangle, Four Gases and Duval pentagon in classifying failure types with various distance metrics namely Canberra, Euclidean, and Bray Curtis. A total of 822 data samples were used to train and validate the model. The results of transformer failure diagnosis show that Duval Triangle, Four Gasses, and Duval Pentagon are the most effective methods in diagnosing transformer failure types. Classification results with the KNN algorithm are strongly influenced by the selection of parameters used to determine each type of class, the highest accuracy in classification with the KNN algorithm is obtained by the Duval triangle method with an accuracy rate of 98,17%.

Actions (login required)

View Item