Ouput-feedback control of combined sewer networks through receding horizon control with moving horizon estimation

An output-feedback control strategy for pollution mitigation in combined sewer networks is presented. The proposed strategy provides means to apply model-based predictive control to large-scale sewer networks, in-spite of the lack of measurements at most of the network sewers. In previous works, the...

Descripción completa

Detalles Bibliográficos
Autores: Joseph Duran, Bernat, Ocampo-Martínez, Carlos|||0000-0001-9251-6044, Cembrano Gennari, Gabriela|||0000-0003-1436-6022
Tipo de recurso: artículo
Fecha de publicación:2015
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/84090
Acceso en línea:https://hdl.handle.net/2117/84090
https://dx.doi.org/10.1002/2014WR016696
Access Level:acceso abierto
Palabra clave:control system synthesis
observability
optimisation
predictive control
output-feedback control
moving horizon estimators
sewer networks
urban water cycle
Classificació INSPEC::Optimisation
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Descripción
Sumario:An output-feedback control strategy for pollution mitigation in combined sewer networks is presented. The proposed strategy provides means to apply model-based predictive control to large-scale sewer networks, in-spite of the lack of measurements at most of the network sewers. In previous works, the authors presented a hybrid linear control-oriented model for sewer networks together with the formulation of Optimal Control Problems (OCP) and State Estimation Problems (SEP). By iteratively solving these problems, preliminary Receding Horizon Control with Moving Horizon Estimation (RHC/MHE) results, based on flow measurements, were also obtained. In this work, the RHC/MHE algorithm has been extended to take into account both flow and water level measurements and the resulting control loop has been extensively simulated to assess the system performance according to different measurement availability scenarios and rain events. All simulations have been carried out using a detailed physically-based model of a real case-study network as virtual reality.