Institutional Repository
Technical University of Crete
Technical University of Crete
EN
|
EL
| URI: | http://purl.tuc.gr/dl/dias/E96C5C1B-9953-42DB-BD0B-1D49269B55B7 | ||
| Year | 2018 | ||
| Type of Item | Diploma Work | ||
| License |
|
||
| Bibliographic Citation | Konstantina Melagoniti, "Optimal operational control of distribution water networks based on predictive control", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2018 https://doi.org/10.26233/heallink.tuc.77311 | ||
| Appears in Collections |
Energy costs for pumping account for the largest part of the total operating cost of water supply networks. In this thesis we deal with pump schedule optimization of water distribution networks based on Model Predictive Control. Model Predictive Control (MPC) is an anticipatory control methodology, that originated in the process industry in the 70’s and has been subject of research for drinking water and wastewater management over the last one or two decades. An MPC controller for a midscale water pumping station is developed in order to optimize the operation plan of the system for minimum energy consumption. An MPC approach is used to determine the optimal operation plan of the pumps, while taking into account physical constraints of the system, such as tank water level lower and upper limits. The MATLAB Simscape block library has been used to build a full model of a midscale pump station in a Simulink modelling and simulation environment. The system consists of four pumps, fed by a reservoir, that pump water to a tank through a pipeline. The initial model has been simulated in order to observe and examine the behavior of the system, and an updated model was designed to introduce the control methodology. The greek pump station of Vlites, in Akrotiri area of Chania (Crete island), is taken as a case study. All detailed information about the operation of this pump station is from OAK AE. The MPC control approach is tested and the results are discussed. Finally, the case study is tested in a 24-hour time horizon and pump operation plans are proposed for different water demand scenarios. The energy costs with the MPC approach are compared to those of empirical pump operation control. The simulations show that MPC has better results regarding energy consumption for certain scenarios and worse for others. When the controller is expected to maintain the tank water level at a very high setpoint, the comparison does not favor the MPC, whereas, when the setpoint is at a lower level, MPC shows an improvement on energy consumption, over the conventional control.
Copyright © DIAS 2013
