Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 44317 (06)
- University: Sharif University of Technology
- Department: Chemical and Petroleum Engineering
- Advisor(s): Shahrokhi, Mohammad
- Abstract:
- Sudden pressure drop and then the gas shortage in city gas stations is one of the most important problems in high-pressure natural gas transmission networks. On the other hand, power consumption of compressors imposes a lot of cost on compressor stations. The conventional schemes to navigate the compressor stations are traditional methods based on experiment and trial and error which cause energy waste and violation of the optimal operating conditions. In this research, two manners of centralized- and decentralized- control are employed in order to navigate compressor stations in specified operating conditions regarding the operational and practical constraints. In this purpose, a sample gas pipeline network is simulated in a commercial software environment as the actual system under study. A linear model is identified based on the data achieved from the actual model. The model obtained is a multivariable system of first order plus dead time transfer functions (FOPD). Regulatory PI controllers and MPC controllers as decentralized and centralized controllers are designed and utilized. The primary results show a high interaction in the plant and then the bad performance of the PI loops in compare to the MPC controller. Hence the interactions of the plant model are reduced using decomposition techniques based on Nyquist Stability Array and Perron-Frobenius interaction assessment. Both static and dynamic compensators are implemented. The mentioned control manners are also applied to control the decoupled model of plant. The results illustrate the better performance of the controllers in the presence of either static or dynamic compensator. However the MPC controller has better performance in comparison to the PI with and without decoupling, which means the better control and the less power consumption. Finally, the Smith predictor extended for MIMO systems and again implemented decentralized controller. The results illustrate the better performance if there is an exact model of the process
- Keywords:
- Pipelines ; Interference ; Compressor Station ; Model Predictive Control ; Centralited Control
Digital Object List
-
محتواي کتاب
- view
Bookmark