PEMODELAN DEBIT ANDALAN SUB DAS CITARIK DENGAN METODE MULTILAYER PERCEPTRON BACKPROPAGATION

BASYIR, RIZAL (2019) PEMODELAN DEBIT ANDALAN SUB DAS CITARIK DENGAN METODE MULTILAYER PERCEPTRON BACKPROPAGATION. S1 thesis, Universitas Mercu Buana Jakarta.

[img]
Preview
Text (HAL COVER)
01_HAL COVER.pdf

Download (28kB) | Preview
[img]
Preview
Text (ABSTRAK)
02_ABSTRAK.pdf

Download (203kB) | Preview
[img]
Preview
Text (LEMBAR PERNYATAAN)
03_SURAT PERNYATAAN.pdf

Download (158kB) | Preview
[img]
Preview
Text (LEMBAR PENGESAHAN)
04_LEMBAR PENGESAHAN.pdf

Download (170kB) | Preview
[img]
Preview
Text (KATA PENGANTAR)
05_KATA PENGANTAR.pdf

Download (71kB) | Preview
[img]
Preview
Text (DAFTAR ISI)
06_DAFTAR ISI.pdf

Download (416kB) | Preview
[img]
Preview
Text (DAFTAR TABEL)
07_DAFTAR TABEL.pdf

Download (204kB) | Preview
[img]
Preview
Text (DAFTAR GAMBAR)
08_DAFTAR GAMBAR.pdf

Download (268kB) | Preview
[img]
Preview
Text (DAFTAR LAMPIRAN)
09_DAFTAR LAMPIRAN.pdf

Download (227kB) | Preview
[img] Text (BAB I)
10_BAB I.pdf
Restricted to Registered users only

Download (90kB)
[img] Text (BAB II)
11_BAB II.pdf
Restricted to Registered users only

Download (557kB)
[img] Text (BAB III)
12_BAB III.pdf
Restricted to Registered users only

Download (122kB)
[img] Text (BAB IV)
13_BAB IV.pdf
Restricted to Registered users only

Download (5MB)
[img] Text (BAB V)
14_BAB V.pdf
Restricted to Registered users only

Download (147kB)
[img] Text (DAFTAR PUSTAKA)
15_DAFTAR PUSTAKA.pdf
Restricted to Registered users only

Download (28kB)
[img] Text (LAMPIRAN)
16_LAMPIRAN.pdf
Restricted to Registered users only

Download (689kB)

Abstract

Hydrological analysis is needed to determine the magnitude of the depandable flow or design depandable flow. One component in the hydrological cycle is rainfall-runoff. Components of rainfall-runoff can be run-off or larger flow such as water flow in the river. In this study reliable discharge modeling using artificial neural networks with backpropagation learning method. Artificial neural network is one of the artificial representations of the human brain that always tries to simulate the learning process in the human brain. The location of the review carried out is in the Citarik Sub-Watershed. The multilayer perceptron ANN with the backpropagation method was used to make the model and verify the model statistically based on the mean square error (MSE), Nash-sutcliffe Efficiency (NSE) and the correlation coefficient (r). The results of the analysis Of the 5 network architecture models that were implemented, the March model 5 gave the most optimum results with MSE = 1.85E-06, and the correlation value in the two efficiency values used showed a correlation of 99%. However, in the testing process the model produces a mean square error of 0.935, and the NSE correlation value = 0.062 and r2 = 0.064. From these results it can be seen that artificial neural networks have sufficient ability to replicate random discharge into artificial models that have almost the same fluctuations and can also be applied in the mainstay discharge modelization even though the testing results are not very accurate because deviations still occur . Keywords : Artificial Neural Network, Multilayer Perceptron, Backpropagation, Depandable Flow, NSE, Citarik Sub-Watershed. Analisa hidrologi diperlukan untuk menentukan besarnya debit andalan atau debit desain. Salah satu komponen dalam siklus hidrologi adalah limpasan hujan. Komponen limpasan hujan dapat berupa run-off (aliran permukaan) ataupun aliran yang lebih besar seperti aliran air di sungai. Pada penelitian ini dilakukan pemodelan debit andalan menggunakan jaringan syaraf tiruan dengan metode pembelajaran backpropagation. Jaringan syaraf tiruan adalah merupakan salah satu representasi buatan dari otak manusia yang selalu mencoba mensimulasikan proses pembelajaran pada otak manusia tersebut. Lokasi tinjauan yang dilakukan adalah di Sub Daerah Aliran Sungai (DAS) Citarik. ANN multi layer perceptron dengan metode back propagation digunakan untuk membuat model serta memverifikasi model tersebut secara statistik berdasarkan nilai mean square error (MSE), Nash-sutcliffe Efficiency (NSE) dan nilai koefisien korelasi (r). Hasil analisis Dari 5 model arsitektur jaringan yang diterapkan, model 5 bulan maret memberikan hasil yang paling optimum dengan MSE= 1.85E-06, dan nilai korelasi pada kedua nilai efisiensi yang digunakan menunjukkan korelasi sebesar 99%. Akan tetapi pada proses pengujian model menghasilkan mean square error sebesar 0.935, dan nilai korelasi NSE = 0.062 dan r2 = 0.064. Dari hasil tersebut terlihat bahwa jaringan syaraf tiruan memiliki kemampuan yang cukup baik dalam mereplikasi debit yang acak ke dalam bentuk model buatan yang memiliki fluktuasi yang hampir sama dan juga dapat diterapkan dalam modelisasi debit andalan walaupun pengujian (testing) hasilnya tidak terlalu akurat karena masih terjadi penyimpangan. Kata Kunci : Jaringan Syaraf Tiruan, Multilayer Perceptron, Backpropagation, Debit Andalan, NSE, Sub DAS Citarik.

Item Type: Thesis (S1)
NIM: 41115010030
Uncontrolled Keywords: Jaringan Syaraf Tiruan, Multilayer Perceptron, Backpropagation, Debit Andalan, NSE, Sub DAS Citarik.
Divisions: Fakultas Teknik > Teknik Sipil
Depositing User: Virda Syifa
Date Deposited: 11 Jul 2019 01:36
Last Modified: 11 Jul 2019 01:36
URI: http://repository.mercubuana.ac.id/id/eprint/49694

Actions (login required)

View Item View Item