Penerapan Algoritma Genetika pada Pelatihan Jaringan Syaraf Tiruan untuk Peramalan Debit Inflow Waduk Sengguruh

Firdaus, DafidEko (2013) Penerapan Algoritma Genetika pada Pelatihan Jaringan Syaraf Tiruan untuk Peramalan Debit Inflow Waduk Sengguruh. Sarjana thesis, Universitas Brawijaya.

Abstract

Waduk merupakan salah satu sumber daya yang memiliki banyak fungsi, diantaranya sebagai sumber energi dalam pengelolaan Pembangkit Listrik Tenaga Air (PLTA), irigasi lahan pertanian, pasokan persediaan Perusahaan Air Minum dan berfungsi sebagai pencegah banjir. Dengan vitalnya peran waduk tersebut diperlukan suatu strategi pengelolaan waduk agar mendapatkan hasil yang optimal dalam pengoperasiannya, salah satunya dengan meramalkan debit inflow (debit masuk). Pada penelitian ini digunakan Jaringan Syaraf Tiruan sebagai model peramalan debit inflow dengan metode pelatihan Algoritma Genetika. Pelatihan Jaringan Syaraf Tiruan menggunakan Algoritma Genetika dilakukan dengan mengkodekan bobot dan bias jaringan kedalam kromosom dan nilai fitness didapat dari error hasil proses feedforward. Pada penelitian ini dapat diketahui parameter genetika mampu mempengaruhi besar nilai fitness diantaranya probabilitas crossover dan jumlah generasi. Bedasarkan hasil uji coba, keakuratan peramalan menggunakan kromosom terbaik mampu mencapai 87,272%.

English Abstract

The reservoir is one resource that has many functions, such as an energy source in the management of Hydroelectric Power Plant (HEPP), irrigation of agricultural land, drinking water supply and serves as a flood prevention. With that many vital role, the reservoir required a management strategy in order to obtain optimal results in the operation, one of them is forecast the inflow. In this study, Artificial Neural Networks are used as inflow forecasting model with Genetic Algorithm training methods. Training Neural Network using Genetic Algorithms is implemented with encode the network weights and biases into the chromosome and fitness values obtained from the cumulative error feedforward process. In this study, several parameters such as the probability of genetic crossover (pc) and the number of generations influence the maximum fitness value obtained in the training process. The results of the training are used to perform forecasting accuracy achieved 87.272%.

Item Type:	Thesis (Sarjana)
Identification Number:	SKR/FTIK/2013/28/051300958
Subjects:	000 Computer science, information and general works > 005 Computer programming, programs, data
Divisions:	Fakultas Ilmu Komputer > Teknik Informatika
Depositing User:	Hasbi
Date Deposited:	28 Mar 2013 10:28
Last Modified:	23 Oct 2021 09:43
URI:	http://repository.ub.ac.id/id/eprint/145815

	Text [0810960038]-SKRIPSI.pdf Download (6MB)
	Text JURNAL-[0810960038].pdf Download (2MB)
	Other (Thumbnails conversion from text to thumbnail_lightbox) lightbox.jpg Download (47kB)
	Other (Thumbnails conversion from text to thumbnail_preview) preview.jpg Download (18kB)
	Other (Thumbnails conversion from text to thumbnail_medium) medium.jpg Download (4kB)
	Other (Thumbnails conversion from text to thumbnail_small) small.jpg Download (1kB)
	Other (Thumbnails conversion from text to thumbnail_lightbox) lightbox.jpg Download (34kB)
	Other (Thumbnails conversion from text to thumbnail_preview) preview.jpg Download (14kB)
	Other (Thumbnails conversion from text to thumbnail_medium) medium.jpg Download (3kB)
	Other (Thumbnails conversion from text to thumbnail_small) small.jpg Download (1kB)

Actions (login required)

View Item