Hadi, Irwin Oktoviantini (2015) Permodelan Kekuatan Beton Menggunakan Data Beton 5 Jam Dengan Metode Adaptive Neuro-Fuzzy Inference System (ANFIS) Matlab. Undergraduate thesis, Institut Technology Sepuluh Nopember.
Preview |
Text
3111100097-Undergraduate Thesis.pdf - Published Version Download (4MB) | Preview |
Abstract
Nasser dan Beaton (1980) telah melakukan percobaan
memprediksi kuat tekan beton umur 28 hari, menggunakan hasil
tes kuat tekan beton 5 jam dengan menggunakan metode
persamaan linear. Tetapi mempunyai nilai error rata-rata (MSE)
sebesar ±11%. Alternatif lain adalah permodelan dengan
Adaptive Neuro-Fuzzy Inference System (ANFIS). Penerapan
ANFIS untuk berbagai aspek permodelan telah banyak dilakukan
dalam sejumlah kajian pada beberapa tahun terakhir dan ANFIS
sangat sesuai untuk permodelan nonlinier. Zhu (2000) telah
menunjukan bahwa ANFIS merupakan metode permodelan tebaik
untuk menganalisi data numerik, karena dalam proses training
didasarkan minimalisasi nilai kesalahan atau root mean square
error (RMSE) dari output-nya. Sehingga ANFIS dapat dijadikan
alternatif untuk memodelkan prediksi kuat tekan beton. Tugas
akhir ini merupakan studi kasus terhadap beberapa data hasil
pengujian oleh Nasser dan Beaton (1980). Studi kasus ini
mencari hubungan variabel dari data mix design dan hasil tes
kekuatan beton 5 jam Nasser dan Beaton (1980) menggunakan
ANFIS sehingga didapat prediksi kekuatan beton umur 28 hari.
Penelitian ini mendapatkan nilai error rata-rata ANFIS sebesar
±7%. Sehingga permodelan menggunakan ANFIS lebih akurat.
===============================================================================================================
Nasser dan Beaton (1980) were experimenting on
prediction of 28 days-old concrete compressive strength using 5
hours concrete compressive strength using regresion method. The
method had ±11% mean square error. Another alternative of
prediction is Adaptive Neuro-Fuzzy Inference System (ANFIS).
ANFIS was used for many studies for years and compatible for
non-linear modelling. Zhu (2000) showed that ANFIS modelling
were the most suitable method for numerical data analysis,
because from data training process based on minimize RMSE, so
ANFIS is suitable for prediction of concrete compressive strength.
This final project was a case study of some experimented data by
Nasser dan Beaton (1980). This case study was looking for data
mix design and concrete compressive strength 5-hours Nasser
dan Beaton (1980) correlation using ANFIS, so generate a 28
days-old concrete compressive strength. This research gets a
±7% mean square error (MSE), which means ANFIS modelling is
more accurate than the previous one.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSS 620.136 Had p |
| Uncontrolled Keywords: | Adaptive Neuro-Fuzzy Inference System (ANFIS); Beton; Prediksi kuat tekan. |
| Subjects: | T Technology > TA Engineering (General). Civil engineering (General) > TA681 Concrete construction |
| Divisions: | Faculty of Civil Engineering and Planning > Civil Engineering > 22201-(S1) Undergraduate Thesis |
| Depositing User: | Mr. Tondo Indra Nyata |
| Date Deposited: | 14 Jun 2019 07:20 |
| Last Modified: | 14 Jun 2019 07:20 |
| URI: | http://repository.its.ac.id/id/eprint/63126 |
Actions (login required)
 |
View Item |