#TITLE_ALTERNATIVE#
In chemical industries, most of binary or multi component mixture separation processes are performed using distillation to produce high quality product. A stable process in the separation process is critical. In order to keep the process stable, a control algorithm is able to maintaining process var...
Saved in:
| Main Author: | |
|---|---|
| Format: | Dissertations |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/10076 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | In chemical industries, most of binary or multi component mixture separation processes are performed using distillation to produce high quality product. A stable process in the separation process is critical. In order to keep the process stable, a control algorithm is able to maintaining process variables at desired values is needed.<p> <br />
<br />
<br />
<br />
<br />
<br />
In this study, a nonlinear predictive control algorithm with optimization using genetic algorithm has been developed and implemented experimentally in a distillation column at Chemical Engineering laboratory, Politeknik Negeri Bandung. The distillation column is able to separating ethanol-water mixture. The optimization of the developed controller successful controls the top column temperature and pressure difference alongside the column constant at desired values.<p> <br />
<br />
<br />
<br />
<br />
<br />
Before the control algorithm was implemented on-line in the distillation column, the dynamic models of the distillation column were derived using identification method and then control algorithm was tested using the identified models through simulation study.<p> <br />
<br />
<br />
<br />
<br />
<br />
The Real time control strategy on the distillation column that has been tested able to keep the top column temperature and pressure difference alongside the column around their each set point, i.e. 76.14 0C which Root Mean Square Error (RMSE = 0.85) and 0.014 bar (which RMSE = 0.00054) respectively.<p> <br />
<br />
<br />
<br />
<br />
<br />
The study gives two contributions. The first contribution is a nonlinear predictive control strategy with optimization using genetic algorithm applied to on-line control of a distillation column. The second one is two nonlinear models of the distillation column in two Multi Input Single Output (MISO) representations, which its parameter estimation was performed using genetic algorithm.<p> <br />
<br />
<br />
<br />
<br />
<br />
It is recommended that this study should be extended to real-time nonlinear predictive control algorithm with optimization using genetic algorithm involving more processes and manipulated variables. In addition, this real-time control can also be extended to Multi Input Multi Output (MIMO) representation with faster computation time. |
|---|