Optimal Control in Water Distribution
Author
Andersen, Benjamin Keldberg
Term
4. term
Education
Publication year
2020
Submitted on
2020-08-20
Abstract
This thesis explores using MPC, a predictive control method that optimizes future actions, to operate a water distribution network. A laboratory setup mimicked a real system with a pump, consumers, pipes, and a water tower. The MPC controller used a cost function designed to lower operating cost and reduce pressure variance. The controller’s behavior was examined first in simulation. Its performance was then evaluated against the two objectives and compared with a simple on/off pump controller. In these evaluations, MPC achieved lower costs and more stable pressure than the on/off controller.
Denne afhandling undersøger brugen af MPC, en forudseende styringsmetode der optimerer fremtidige handlinger, til styring af et vanddistributionsnet. Et laboratorieanlæg efterlignede et rigtigt system med pumpe, forbrugere, rør og et vandtårn. MPC-controllerens målfunktion blev udformet til at sænke driftsomkostninger og reducere trykvariation. Controllerens adfærd blev først undersøgt i simulation. Ydelsen blev derefter vurderet i forhold til de to mål og sammenlignet med en simpel tænd/sluk-controller til pumpen. I disse evalueringer gav MPC lavere omkostninger og mere stabilt tryk end tænd/sluk-controlleren.
[This apstract has been rewritten with the help of AI based on the project's original abstract]
Keywords