Nonlinear predictive control for the concentrations profile regulation under unknown reaction disturbances in a fuel cell anode gas channel
In this work, a nonlinear model predictive control (NMPC) strategy is proposed to regulate the concentrations of the different gas species inside a Proton Exchange Membrane Fuel Cell (PEMFC) anode gas channel. The purpose of the regulation relies on the rejection of the unmeasurable perturbations th...
| Autores: | , , |
|---|---|
| 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:2117/27120 |
| Acceso en línea: | https://hdl.handle.net/2117/27120 https://dx.doi.org/10.1016/j.jpowsour.2015.02.033 |
| Access Level: | acceso abierto |
| Palabra clave: | PEM fuel cell systems control nonlinearities observability power system control predictive control. Concentration regulation distributed model NMPC fuel cell control fuel cell observation Piles de combustible -- Control electrònic Àrees temàtiques de la UPC::Informàtica::Automàtica i control |
| Sumario: | In this work, a nonlinear model predictive control (NMPC) strategy is proposed to regulate the concentrations of the different gas species inside a Proton Exchange Membrane Fuel Cell (PEMFC) anode gas channel. The purpose of the regulation relies on the rejection of the unmeasurable perturbations that affect the system: the hydrogen reaction and water transport terms. The model of the anode channel is derived from the discretisation of the partial differential equations that define the nonlinear dynamics of the system, taking into account spatial variations along the channel. Forward and backward discretisations of the distributed model are employed to take advantage of the boundary conditions of the problem. A linear observer is designed and implemented to perform output-feedback control of the plant. This information is fed to the controller to regulate the states towards their desired values. Simulation results are presented to show the performance of the proposed control method over a given case study. Different cost functions are compared and the one with minimum state-regulation error is identified. Suitable dynamic responses are obtained facing the different considered disturbances. |
|---|