Wiryawan, Gede Panji (2016) Prediksi Sifat Termal Bata Ringan Berdasarkan Variasi Komposisi Menggunakan Jaringan Syaraf Tiruan. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
2412100065-undergraduate theses.pdf - Published Version Download (2MB) | Preview |
Preview |
Text
2412100065-paperpdf.pdf - Published Version Download (793kB) | Preview |
Preview |
Text
2412100065-presentationpdf.pdf - Presentation Download (2MB) | Preview |
Abstract
Pada penelitian Tugas Akhir ini, Jaringan Syaraf Tiruan (JST)
Backpropagation digunakan untuk memprediksi konduktivitas termal
bata ringan jenis Autoclaved Aerated Concrete (AAC). Berdasarkan
pelatihan dan evaluasi yang telah dilakukan terhadap 10 model JST
dengan jumlah hidden node 1 sampai 10, didapati bahwa JST dengan 3
hidden node memiliki performance yang paling baik. Hal ini diketahui
dari nilai MSE (Mean Square Error) rata-rata validasi untuk tiga kali
pelatihan sebesar 0,003269. Jaringan ini selanjutnya digunakan untuk
memprediksi konduktivitas termal empat jenis bata ringan. Hasil prediksi
untuk masing-masing bata ringan AAC-1, AAC-2, AAC-3 dan AAC-4
berturut-turut adalah sebesar 0,243 W/mK; 0,29 W/mK; 0,32 W/mK; dan
0,32 W/mK. Selanjutnya, JST digunakan untuk mengetahui pengaruh
komposisi Silikon (Si), Kalsium (Ca), Aluminium (Al), dan massa jenis
terhadap konduktivitas bata ringan. Hasil simulasi JST menunjukkan
bahwa konduktivitas termal meningkat seiring dengan meningkatnya
komposisi Si dan massa jenis, namun terjadi penurunan konduktivitas
termal seiring kenaikan komposisi Al. ========== In this study, backpropagation neural network was used to predict
thermal conductivity of autoclaved aerated concrete (AAC). Ten network
models with various number of hidden nodes (1 to 10 hidden nodes) were
trained each three times. Among the ten network models, network with 3
hidden nodes came out with the best performance (average MSE
validation = 0,003269). This network was used to predict thermal
conductivity of four AAC samples; i.e. AAC-1, AAC-2, AAC-3, AAC-4 and
gave the following results: AAC-1 = 0.243 W/mK, AAC-2 = 0.29 W/mK,
AAC-3 = 0.32 W/mK, AAC-4 = 0.32 W/mK. The network was then used
to examine the influence of composition variation of Calcium (Ca),
Aluminium (Al), Silicone (Si) and density on thermal conductivity of AAC.
It was found that thermal conductivity increase along with density and
concentration of Si, while the increase in concentration of Al leads to the
decrease of thermal conductivity of AAC.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSF 620.118 Wir p 3100017068968 |
| Uncontrolled Keywords: | Jaringan Syaraf Tiruan (JST) backpropagation, bata ringan Autoclaved Aerated Concrete (AAC), konduktivitas termal, Backpropagation neural network, autoclaved aerated concrete, thermal conductivity |
| Subjects: | T Technology > TA Engineering (General). Civil engineering (General) > TA418.9 Composite materials. Laminated materials. |
| Divisions: | Faculty of Industrial Technology > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | - Davi Wah |
| Date Deposited: | 21 Oct 2019 03:58 |
| Last Modified: | 21 Oct 2019 03:58 |
| URI: | http://repository.its.ac.id/id/eprint/71226 |
Actions (login required)
 |
View Item |