Publication
Multi-layer Model Predictive Control of Complex Water Systems
PhD Thesis (2015)
Month
11
Director
Gabriela Cembrano Gennari
Director
Vicenç Puig Cayuela
File
Authors
Abstract
This thesis is devoted to design a multi-layer MPC controller applied to the complex water network taking into account that the different layers with different time scales and control objectives have their own controller. A two-layer temporal hierarchy coordinating scheme has been applied to coordinate the MPC controllers for the supply and transportation layers. An integrated real-time simulation-optimization approach which contributes to consider the effect of more complex dynamics, better represented by the simulation model, has been developed for regional water networks. The use of the combined approach of optimization and simulation coordination between simulator and optimizer allows to test the proposed multi-layer MPC in a feedback scheme using a realistic simulator of the regional network.
The second part of this thesis is focused on the design of a control scheme which uses the combination of linear MPC with a constraint satisfaction problem (CSP) to optimize the non-linear operational control of DWNs. The methodology has been divided into two functional layers: First, a CSP algorithm is used to transfer non-linear DWN pressure equations into linear constraints, which can enclose the feasible solution set of the hydraulic non-linear problem during the optimizing process. The network aggregation method (NAM) is used to simplify a complex water network into an equivalent conceptual one for the bidirectional network before the use of CSP. Then, a linear MPC with added linear constraints is solved to generate optimal control strategies which optimize the control objective. The proposed approach is simulated using Epanet to represent the real DWN. Non-linear MPC is used for validation using a generic operational tool for controlling water networks named PLIO.
A two-layer scheduling scheme for pump stations in a water distribution network has also been designed in the second part of this thesis. The upper layer, which works in one-hour sampling time, uses MPC to produce continuous flow set-points for the lower layer. While in the lower layer, a scheduling algorithm has been used to translate the continuous flow set-points to a discrete (ON-OFF) control operation sequence of the pump stations with the constraints that pump stations should draw the same amount of water as the continuous flow set-points provided by the upper layer. The tuning parameters of such algorithm are the lower layer control sampling period and the number of parallel pumps in the pump station.
Categories
automation, control theory, optimisation.
Author keywords
Congcong Sun
Follow us!