Ghina, Firda Fikriyatul (2024) Klasifikasi Gangguan Transmisi Tegangan Tinggi 150 kV Menggunakan Metode Artificial Neural Network (ANN). Diploma thesis, Institut Teknologi Sepuluh Nopember.
![]() |
Text
2040201156_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (8MB) | Request a copy |
Abstract
Energi listrik memegang peran penting dalam mendukung kebutuhan hidup, dan permintaan terus meningkat seiring berjalannya waktu. Sistem transmisi tenaga listrik bertujuan untuk mengalirkan listrik dari sumber pembangkit ke konsumen. Namun, dalam proses tersebut, gangguan pada transmisi dapat timbul karena berbagai faktor, termasuk gangguan pada peralatan, intervensi manusia, dan faktor alam. Gangguan semacam ini dapat menyebabkan terputusnya aliran listrik dan menyebabkan pemadaman, yang berdampak pada ketidaknyamanan dan bahkan kerugian bagi pengguna listrik. Proyek Akhir ini, gangguan pada transmisi dikelompokkan menjadi empat kelas utama: gangguan akibat pohon, gangguan oleh hewan, gangguan petir, dan konduktor putus. Untuk mencegah gangguan dan kesalahan deteksi penyebab gangguan, pemantauan saluran transmisi dianggap sebagai upaya yang efektif. Ada berbagai metode pemantauan dan salah satunya adalah Artificial Neural Network (ANN), ANN adalah sebuah metode pembelajaran mesin yang efektif pada beberapa bidang termasuk klasifikasi data, regresi, dan prediksi. Dalam Proyek Akhir ini ANN diterapkan untuk mengklasifikasi gangguan yang dihasilkan oleh alat Disturbance Fault Recorder (DFR). Karakteristik tiap gelombang gangguan bervariasi, mencakup frekuensi, amplitudo, dan durasi. Data tersebut dianalisis menggunakan Fast Fourier Transform (FFT) untuk mengubah data dari domain waktu ke domain frekuensi. Analisis FFT ini memudahkan klasifikasi pola frekuensi yang terkait dengan jenis gangguan tertentu. Setelah dianalisis dilanjutkan dengan proses klasifikasi. Klasifikasi menggunakan ANN meliputi beberapa proses seperti normalisasi data, feed-forward dan backpropagation. Pada penelitian dilakukan 30 kali percobaan untuk menentukan arsitektur ANN. Berdasarkan penelitian, arsitektur terbaik menggunakan 2 hidden layer dengan 10 neuron di hidden layer 1 dan 8 neuron di hidden layer 2, dengan batch size sebesar 40 dan iterasi sebanyak 200 kali, yang mendapatkan nilia loss sebesar 0,0166 dan akurasi sebesar 99,36%.
=================================================================================================================================
Electrical energy holds an important role in supporting the needs of life, and demand continues to increase over time. The power transmission system aims to deliver electricity from generating sources to consumers. However, in the process, disruptions in transmission can arise due to various factors, including equipment interference, human intervention, and natural factors. Such disturbances can lead to power interruptions and cause blackouts, which result in inconvenience and even losses for electricity users. In this final project, transmission faults are classified into four main classes: tree faults, animal faults, lightning faults, and conductor breaks. Transmission line monitoring is considered an effective endeavor to prevent disturbances and misdetection of the cause of the disturbance. There are various monitoring methods and one of them is Artificial Neural Network (ANN), ANN is a machine learning method that is effective in several fields including data classification, regression, and prediction. In this final project, ANN is applied to classify the fault patterns generated by the Disturbance Fault Recorder (DFR). The characteristics of each fault wave vary, including frequency, amplitude, and duration. The data is analyzed using Fast Fourier Transform (FFT) to transform the data from time domain to frequency domain. This FFT analysis facilitates the classification of frequency patterns associated with certain types of disturbances. After the analysis, the classification process is continued. Classification using ANN includes several processes such as data normalization, feed-forward and backpropagation. The research conducted 30 experiments to determine the ANN architecture. Based on the research, the best architecture uses 2 hidden layers with 10 neurons in hidden layer 1 and 8 neurons in hidden layer 2, with a batch size of 40 and 200 iterations, which gets a loss value of 0.0166 and accuracy of 99.36%.
Item Type: | Thesis (Diploma) |
---|---|
Uncontrolled Keywords: | Artificial Neural Network, Disturbance Fault Recorder, Fast Fourier Transform, Gangguan, Sistem Transmisi, Transmission System, Disturbance |
Subjects: | Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA404 Fourier series Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > T Technology (General) > T174 Technological forecasting T Technology > T Technology (General) > T57.5 Data Processing T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK1007 Electric power systems control |
Divisions: | Faculty of Vocational > 36304-Automation Electronic Engineering |
Depositing User: | Firda Fikriyatul Ghina |
Date Deposited: | 21 Aug 2024 03:33 |
Last Modified: | 21 Aug 2024 03:33 |
URI: | http://repository.its.ac.id/id/eprint/115482 |
Actions (login required)
![]() |
View Item |