Taming Heterogeneity and Complexity of Embedded Control

Hybrid Model Predictive Control Applied on Sewer Networks

The Barcelona Case Study

Carlos Ocampo-Martínez — Ari Ingimundarson — Vicenç Puig — Joseba Quevedo

Automatic Control DepartmentTechnical University of Catalonia (UPC)Rambla de Sant Nebridi 10, 08222,Terrassa, Spain{carlos.ocampo,ari.ingimundarson,vicenc.puig,joseba.quevedo}@upc.edu

ABSTRACT. In this paper, hybrid model predictive control (HMPC) for sewer networks is introduced. A hybrid model of the sewage system is presented which allows capturing inherent phenomena such as tank and collector overflows that are not always present but depend on the system state. The model is used for predictive control to reduce combined sewer overflows (CSO) and pollution while the sewage treatment is improved. A simulation study is presented where the hybrid control is designed and analyzed for real rain scenarios.

KEYWORDS: Model predictive control, hybrid systems, hybrid control, sewage systems, urban drainage

1. Introduction

Under the city of Barcelona there is a complex sewer network, almost 1500 Km in length, which is characterized for being unitary, that is, its collectors carry together rain water and residual water. Having a unitary system, it is specially important that the sewer network is capable of complying with its role during severe rains as well as in non-threatening weather, see [CEM 04]. The yearly rainfall in Barcelona is not very high (600 mm/year), but it includes heavy storms typical of the Mediterranean climate ...