PERBANDINGAN PERAMALAN BEBAN PUNCAK BERBASIS ALGORITMA FUZZY C-MEANS DAN FUZZY SUBTRACTIVE CLUSTERING DENGAN OPTIMASI CLUSTER HARI DAN JUMLAH INPUT

Yunus, M., Muhammad (2017) PERBANDINGAN PERAMALAN BEBAN PUNCAK BERBASIS ALGORITMA FUZZY C-MEANS DAN FUZZY SUBTRACTIVE CLUSTERING DENGAN OPTIMASI CLUSTER HARI DAN JUMLAH INPUT. S1 thesis, Universitas Pendidikan Indonesia.

[img] Text
S_TE_1006786_Title.pdf

Download (269kB)
[img] Text
S_TE_1006786_Abstract.pdf

Download (137kB)
[img] Text
S_TE_1006786_Table_Of_Content.pdf

Download (165kB)
[img] Text
S_TE_1006786_Chapter1.pdf

Download (262kB)
[img] Text
S_TE_1006786_Chapter2.pdf
Restricted to Staf Perpustakaan

Download (846kB)
[img] Text
S_TE_1006786_Chapter3.pdf

Download (926kB)
[img] Text
S_TE_1006786_Chapter4.pdf
Restricted to Staf Perpustakaan

Download (2MB)
[img] Text
S_TE_1006786_Chapter5.pdf

Download (138kB)
[img] Text
S_TE_1006786_Bibliography.pdf

Download (190kB)
[img] Text
S_TE_1006786_Appendix.pdf
Restricted to Staf Perpustakaan

Download (1MB)
Official URL: http://repository.upi.edu

Abstract

Peramalan beban listrik merupakan hal yang sangat penting dalam perencanaan dan oprerasi sistem tenaga yang efektif. Berbagai metode digunakan dalam peramalan beban listrik, terutama metode berbasis kecerdasan buatan yang sekarang banyak dikembangkan dengan berbagai model perhitungan sistem maupun penentuan model input. Penelitian ini mengkaji tentang peramalan beban listrik jangka pendek dengan metode berbasis algoritma Fuzzy C-Means (FCM) dan Fuzzy Subtractive Clustering (FSC). Hasil peramalan beban listrik kemudian dibandingkan dengan hasil peramalan yang telah dilakukan PT. PLN (Persero) yang digunakan dalam Rencana Beban Sistem PLN (RBS PLN) P3B Jawa Barat. RBS PLN pada data target yang digunakan dalam penelitian ini menunjukan kesalahan sebesar 13,21%, 8,15% dan 10,42% berturut-turut pada klasifikasi hari kerja, akhir pekan dan libur nasional. Nilai tersebut dinilai terlalu tinggi dibandingkan dengan laporan peramalan beban listrik yang sekarang banyak dikembangkan. Tingkat akurasi dengan metode clustering sangat dipengaruhi oleh klasifikasi input. Pada penelitian ini input peramalan merupakan data aktual beban listrik sejak tahun 2006 s.d 2014 dengan optimasi banyaknya input sample dan cluster hari berdasarkan hari kerja, akhir pekan dan hari libur nasional yang dihitung dengan menggunakan user interface. Hasil perhitungan menunjukan peramalan yang lebih akurat daripada RBS PLN. Nilai indeks MAPE hasil peramalan beban listrik dengan metode berbasis algortima FCM adalah 0,02%, 0,13% dan 0,07%. Sedangkan dengan algoritma FSC ada 0,18%, 0,88% dan 0,29% berturut-turut pada cluster hari kerja, akhir pekan dan hari libur nasional.;--- Electricity load forecasting is very important in the planning and operation of an effective power system. Various methods are used in electrical load forecasting, especially artificial intelligent based methods that are now widely developed with various models of system calculation and input model determination. This study examines the short-term load forecasting using Fuzzy C-Means (FCM) and Fuzzy Subtractive Clustering (FSC) -based algorithms. Results of electricity load forecasting then compared with forecasting results that have been done by PT. PLN (Persero) used in PLN Load System Plan (RBS PLN) P3B West Java. RBS PLN on the target data used in this study shows errors of 13.21%, 8.15% and 10.42% respectively on the classification of weekdays, weekends and national holidays. The value is considered too high compared to the report of electricity load forecasting that is now widely developed. The level of accuracy with the clustering method is strongly influenced by the classification of inputs. In this study input sample is the actual data electrical load since 2006 - 2014 with the optimization of the number of input and cluster days based on weekdays, weekends and national holidays that are calculated using a user interface. The calculation results show more accurate forecasting than RBS PLN. The MAPE index value of electricity load forecasting results with FCM algortihm based method is 0.02%, 0.13% and 0.07%. While the FSC algorithm is 0.18%, 0.88% and 0.29% respectively on the cluster of weekdays, weekends and national holidays.

Item Type: Skripsi,Tesis,Disertasi (S1)
Additional Information: No. Panggil: S TE YUN p-2017; Pembimbing: I. Ade Gafar Abdullah , II. Hasbullah; NIM: 1006786.
Uncontrolled Keywords: peramalan beban listrik jangka pendek, cluster, fuzzy c-means, fuzzy subtractive clustering, user interface, MAPE, short term load forecasting, cluster, fuzzy c-means, fuzzy subtractive clustering, user interface, MAPE.
Subjects: T Technology > T Technology (General)
T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions: Fakultas Pendidikan Teknologi dan Kejuruan > Jurusan Pendidikan Teknik Elektro
Depositing User: DAM staf
Date Deposited: 30 Nov 2018 02:44
Last Modified: 30 Nov 2018 02:44
URI: http://repository.upi.edu/id/eprint/32241

Actions (login required)

View Item View Item