Aris, Sofidul (2010) Perancangan Sistem Pengendalian Level Dengan Jaringan Syaraf Tiruan Pada Steam Drum Boiler B-1102 Di PT. Petrokimia Gresik. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
2404100076-Undergraduate_Theses.pdf - Accepted Version Download (2MB) | Preview |
Preview |
Text
2404100076-Paper.pdf - Accepted Version Download (615kB) | Preview |
Preview |
Text
2404100076-Presentation.pdf - Presentation Download (437kB) | Preview |
Abstract
Pengendalian level yang selama ini digunakan pada steam
drum adalah dengan pengendali PID. Pengendali konvensional
ini dirasa kurang mampu memberikan perfomansi yang baik
karena sifatnya yang linier padahal proses-proses pada industri
biasanya nonlinear, multivariabel dan komplek. Diperlukan suatu
pengendalian yang mampu menangani sifat kenonlinearan steam
drum tersebut. Dengan kemampuan belajar terhadap berbagai
perubahan kondisi yang tak terduga maka jaringan syaraf tiruan
digunakan untuk merancang sistem pengendalian level dengan
struktur model NNARX (Neural Network AutoRegresive with
eXternal input), algoritma pembelajaran Levenberg Marquardt
dan Direct Inverse Control berbasis jaringan syaraf tiruan.
Penelitian dilakukan dengan memodelkan plant serta
pengendali sehingga dihasilkan RMSE = 0.0019 untuk model
plant dan RMSE 0.0014 untuk model pengendali. Didapat
performansi pengendalian DIC dan PID berbasis JST, level
steam drum yakni; DIC menghasilkan Mp = 2.24%, ts = 30
sekon, dan pengendali PID menghasilkan Mp = 2.26%, ts =
31 sekon, pada set point 75 % (4.533meter) level steam
drum. ========== The PID controller has been used in the steam drum level
control for many years . This conventional controller is less able
to provide good performance because of it is linearity whereas
industrial processes are usually nonlinear, multivariable and
complex. A control that can handle steam drum unlinearity
properties is needed. With the ability to learn to various
unforeseen circumstances changes, the neural network is used to
design the control system with structure level NNARX model
(Neural Network AutoRegresive with external input), Levenberg
Marquardt learning algorithm and the Direct Inverse Control
based on artificial neural networks. Research done by modeling
the plant and controller so that the resulting RMSE = 0.0019
for the model plant and RMSE 0.0014 for controller model.
The performance gained control by DIC and PID
controllers using neural network, of the level steam drum;
DIC yield Mp = 2.24%, ts = 30 seconds, and PID
controllers yield Mp = 2.26%, ts = 31 seconds at set point
75% (4.533 meters) level steam drum.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSF 006.3 Ari p 3100010040826 |
| Uncontrolled Keywords: | Steam Drum, Pengendalian level, Jaringan Syaraf Tiruan, Direct Inverse Control, Level control, Neural Network |
| Subjects: | Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) |
| Divisions: | Faculty of Industrial Technology > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | - Davi Wah |
| Date Deposited: | 27 Nov 2019 08:11 |
| Last Modified: | 27 Nov 2019 08:11 |
| URI: | http://repository.its.ac.id/id/eprint/72089 |
Actions (login required)
 |
View Item |