Sofyan, Willy Wigia (2014) ESTIMASI BEBAN PUNCAK HARIAN BERDASARKAN KLUSTER TIPE HARI BERBASIS ALGORITMA HYBRID SWARM PARTICLE-ARTIFICIAL NEURAL NETWORK. S1 thesis, Universitas Pendidikan Indonesia.
|
Text
S_PE_1000141_Title.pdf Download (100kB) | Preview |
|
|
Text
S_PE_1000141_Table_Of_Content.pdf Download (203kB) | Preview |
|
|
Text
S_PE_1000141_Abstract.pdf Download (199kB) | Preview |
|
|
Text
S_PE_1000141_Chapter1.pdf Download (113kB) | Preview |
|
![[img]](http://repository.upi.edu/style/images/fileicons/text.png) |
Text
S_PE_1000141_Chapter2.pdf Restricted to Staf Perpustakaan Download (434kB) | Request a copy |
|
|
Text
S_PE_1000141_Chapter3.pdf Download (652kB) | Preview |
|
|
Text
S_PE_1000141_Chapter4.pdf Restricted to Staf Perpustakaan Download (603kB) | Request a copy |
|
|
Text
S_PE_1000141_Chapter5.pdf Download (43kB) | Preview |
|
|
Text
S_PE_1000141_Bibliography.pdf Download (83kB) | Preview |
|
|
Text
S_PE_1000141_Appendix.pdf Restricted to Staf Perpustakaan Download (300kB) | Request a copy |
Abstract
Prediksi beban listrik jangka pendek merupakan salah satu perencanaan operasi sistem tenaga listrik yang memiliki peranan penting dalam hal mewujudkan operasi yang ekonomis. Hasil prediksi tersebut dapat dijadikan masukan utama dalam unit commitment, economic dispatch, ataupun studi aliran daya. Skripsi penelitian ini bertujuan untuk melakukan studi penggunaan algoritma hibrid PSO-JST dalam melakukan prediksi beban puncak harian jangka pendek dengan membuat kluster data berdasarkan tipe hari yang berbeda. Data historis menggunakan data pengeluaran beban listrik dari P3B PT.PLN (Persero) Area III Jawa Barat UPB-Cigereleng. Perhitungan dilakukan dengan menggunakan perangkat lunak MATLAB untuk mengetahui tingkat akurasi prediksi dan nilai MAPE (Mean Absolute Percentage Error) pada algoritma HPSO-JST dibandingkan dengan data Rencana Beban Sistem (RBS) PT.PLN dan algoritma backpropagation-jaringan syaraf tiruan (BP-JST) tanpa dikombinasikan dengan algoritma particle swarm optimization (PSO). Hasil simulasi menunjukkan bahwa hasil prediksi beban puncak berbasis algoritma HPSO-JST memberikan tingkat akurasi yang baik dan nilai MAPE yang kecil serta stabil dibawah 2%, jika dibandingkan dengan prediksi RBS-PLN dan BP-JST. Hasil prediksi beban yang akurat akan menghasilkan efisiensi kepada perusahaan listrik sehingga dapat menekan biaya operasional pembangkitan dan tentunya secara tidak langsung akan berdampak pada murahnya biaya produksi listrik. Kata Kunci : Prediksi Beban Puncak Jangka Pendek, Tipe Hari, Algoritma Hybrid Particle Swarm Optimization Jaringan Syaraf Tiruan, Mean Absolute Percentage Error. Short-term electrical load prediction is one of the operation planning of electric power system has an important role in terms of realizing the economical operation. The prediction results can be used as a major input in unit commitment, economic dispatch, or load flow studies. Final research aims to study the use of hybrid PSO-ANN algorithm to predict the short-term daily peak loads by creating clusters of data based on different types of days. Historical data using expenditure data from the electrical load PT PLN P3B (Persero) Area III West Java UPB-Cigereleng. Calculations were performed using MATLAB to determine the level of accuracy of prediction and the value of MAPE (Mean Absolute Percentage Error) HPSO-ANN algorithm compared with the data Rencana Beban Sistem (RBS) PT.PLN and algorithm-back propagation neural network (BP-ANN) without algorithm combined with particle swarm optimization (PSO). The simulation results prove that the results predicted peak load HPSO algorithm-based ANN gave a good degree of accuracy and MAPE values are small and stable below 2%, when compared with the RBS-PLN predictions and BP-ANN. The results are accurate load prediction will result in efficiencies to the electric company so as to reduce the operational costs of generation and certainly will indirectly have an impact on the low cost of electricity production. Keywords : Prediction of Short-Term Peak Load, Type Day, Hybrid Particle Swarm Optimization Algorithm Neural Network, Mean Absolute Percentage Error.
| Item Type: | Thesis (S1) |
|---|---|
| Additional Information: | No.Pangil : S PE SOF e-2014 |
| Uncontrolled Keywords: | Prediksi Beban Puncak Jangka Pendek, Tipe Hari, Algoritma Hybrid Particle Swarm Optimization Jaringan Syaraf Tiruan, Mean Absolute Percentage Error. |
| Subjects: | L Education > L Education (General) |
| Divisions: | Fakultas Pendidikan Teknologi dan Kejuruan > Jurusan Pendidikan Teknik Elektro |
| Depositing User: | DAM STAF Editor |
| Date Deposited: | 27 Feb 2015 06:53 |
| Last Modified: | 27 Feb 2015 06:53 |
| URI: | http://repository.upi.edu/id/eprint/13543 |
Actions (login required)
 |
View Item |