A K-NN Based Model for Predicting Electrical Grid Failures from Maintenance Records

Devina Harfia, Mahdi Mahdi, Umri Erdiansyah

Abstract


Disturbances in distribution networks can cause outages and reduce system reliability. However, PLN’s routine maintenance is not yet supported by predictive systems that optimally utilize historical data to identify potential risks. This study aims to develop a risk classification system for distribution network disturbances using the K-Nearest Neighbor (K-NN) method. The data were collected from substation and Medium Voltage Overhead Line (SUTM) maintenance, then processed through preprocessing, risk labeling using clustering, and classification with K-NN. The system was tested on 13 feeders, resulting in 8 categorized as LOW risk, 3 as MEDIUM, and 2 as HIGH. The parameter K was determined through cross-validation, where K = 5 was selected as the optimal value. Evaluation results showed an accuracy of 83.75%, with the best performance obtained in the LOW class (precision, recall, and F1-score of 93.41%). The HIGH class achieved 82.76% precision and 76.19% F1-score, while the MEDIUM class was relatively lower with 62.5% precision and 66.67% F1-score. Overall, the model achieved an average precision of 79.6%, recall of 78.5%, and F1-score of 78.8%. These findings indicate that the K-NN method is fairly effective in predicting distribution network disturbance risks and can serve as a decision-support tool for preventive maintenance in PLN’s operations.

References


K. Octavianus Bachri, E. Jonathan, and A. D. Soewono, “Analisis Gangguan Pada Sistem Gardu Listrik Berbasis Gas Insulated Switchgear 150 Kv Dengan Metode High Voltage Test,†Cylind. J. Ilm. Tek. Mesin, vol. 9, pp. 2252–925X, 2023.

T. Andriani, M. Hidayatullah, S. Esabella, and P. Studi Teknik Elektro, “Pemeliharaan Jaringan Distribusi di PT. PLN ULP 2 Mawasangka,†Hexagon, vol. 2, no. 1, 2021.

S. Amalia and E. Saputra, “Pemeliharaan Jaringan Saluran Udara Tegangan Menengah (SUTM ) 20 kV Feeder Mata Air,†J. Tek. Elektro Inst. Teknol. Padang, vol. 9, no. 2, 2020, doi: 10.21063/JTE.2020.3133911.

S. Arena, G. Manca, S. Murru, P. F. Orrù, R. Perna, and D. Reforgiato Recupero, “Data Science Application for Failure Data Management and Failure Prediction in the Oil and Gas Industry: A Case Study,†Appl. Sci., vol. 12, no. 20, Oct. 2022, doi: 10.3390/app122010617.

S. Rezig, Z. Achour, and N. Rezg, “RETRACTED: Using Data Mining Methods for Predicting Sequential Maintenance Activities,†Appl. Sci., vol.8, no. 11, Nov. 2018, doi: 10.3390/app8112184.

M. Kafil, “Penerapan Metode K-Nearest Neighbor Untuk Prediksi Penjualan Berbasis Web Pada Boutiq Dealove Bondowoso,†2019.

L. Abd, R. Hakim, A. A. Rizal, and D. Ratnasari, “Aplikasi Prediksi Kelulusan Mahasiswa Berbasis K-Nearest Neighbor (K-NN),†2019.

Siska Ayu Widiana, Iqbal Firdaus, Edwin Tenda, and Eliasta Ketaren, “Sistem Informasi Prediksi Penjualan Produk Thrift Di Toko Manado Menggunakan Algoritma K-Nearest Neighbor (K-NN),†J. TIMES, vol. 12, no. 2, pp. 52–57, 2023, doi: 10.51351/jtm.12.2.2023708.

Roziana, E. Ramadhani, and S. Anwar, “Analisis Gangguan Listrik Di Pltu Nagan Raya Menggunakan Diagram Kendali Exponentially Weighted Moving Average Dan Decision On Belief,†INTERVAL J. Ilm. Mat., vol. 2, no. 2, pp. 104–117, 2022.

W. Pratama et al., “Prediksi Pemeliharaan Transformator Distribusi Berbasis Artificial Neural Network,†SNESTIK, p. 112, 2023, doi: 10.31284/p.snestik.2023.4004.

M. Yunus and N. K. A. Pratiwi, “Prediksi Status Gizi Balita Dengan Algoritma K-Nearest Neighbor (K-NN) di Puskemas Cakranegara,†JTIM J. Teknol. Inf. dan Multimed., vol. 4, no. 4, pp. 221–231, Feb. 2023, doi: 10.35746/jtim.v4i4.328.

N. Kwintarini Suparman, B. Arif Dermawan, and T. Nur Padilah, “Prediksi Barang Keluar TB. Wijaya Bangunan Menggunakan Algoritma K-NN Regression dengan RStudio,†MEI, 2021.

Rachmadhany Iman, Basuki Rahmat, and Achmad Junaidi, “Implementasi Algoritma K-Means dan Knearest Neighbors (K-NN) Untuk Identifikasi Penyakit Tuberkulosis Pada Paru-Paru,†Repeater Publ. Tek. Inform. dan Jar., vol. 2, no. 3, pp. 12–25, 2024, doi: 10.62951/repeater.v2i3.77.

G. Dimas Prenata, “Klasifikasi Keandalan Sistim Distribusi Tenaga Listrik Di Pt. Pln (Persero) Up3 Surabaya Selatan Menggunakan Metode K-Nearest Neighbor (K-NN),†Jl. Semolowaru No, vol. 11, no. 3, p. 60118, 2023, doi:10.23960/jitet.v11i3%20s1.3397.

M. A. Harriz and H. Setiyowati, “Komparasi Algoritma Decision Tree Dan K-NN Dalam Mengklasifikasi Daerah Berdasarkan Produksi Listrik,†JIKO (Jurnal Inform. dan Komputer), vol. 7, no. 2, p. 167, 2023, doi: 10.26798/jiko.v7i2.787.

Retno Wahyusari, “Perbandingan K-Nearest Neighbor (K-NN) Dan Support Vector Regression (Svr) Untuk Prediksi Konsumsi Energi Listrik,†Pros. Snast, no. November, pp. G44-50, 2024, doi: 10.34151/prosidingsnast.v1i1.5034.

P. Hidayat, R. Kurniawan, Y. A. Wijaya, and T. Suprapti, “Optimasi Algoritma K-Nearest (K-NN) Neighbors Pada Prediksi Risiko Penyakit Kardiovaskular,†J. Inform. dan Tek. Elektro Terap., vol. 13, no. 1, 2025, doi: 10.23960/jitet.v13i1.5864.

R. Joto and M. Urfan Barran Rusyda Marzuq, “Analisis Perencanaan Sistem Jaringan Distribusi Listrik dan Perkembangan Beban Pada Perumahan The Grand Kenjeran Surabaya,†ELPOSYS J. Sist. Kelistrikan, vol. 09, no. 3, 2022.

R. J. Alfirdausy and S. Bahri, “Implementasi Algoritma K-Nearest Neighbor untuk Klasifikasi Diagnosis Penyakit Alzheimer,†Techno.Com, vol. 22, no.3, pp. 635–642, 2023, doi: 10.33633/tc.v22i3.8393.

A. Putri, C. Syaficha Hardiana, E. Novfuja, F. Try Puspa Siregar, Y. Fatma, and R. Wahyuni, “Komparasi Algoritma K-NN, Naive Bayes dan SVM untuk Prediksi Kelulusan Mahasiswa Tingkat Akhir,†MALCOM Indones. J. Mach. Learn. Comput. Sci. J. Homepage, vol. 3, no. 1, pp. 20–26, 2023, [Online]. Available: https://journal.irpi.or.id/index.php/malcom/article/view/610


Refbacks

  • There are currently no refbacks.


Indexing :

Creative Commons License
Journal of Informatics Engineering and Software Applications (JIEngS) licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.