Efficient experimental energy management operating for FC/battery/SC vehicles via hybrid Artificial Neural Networks-Passivity Based Control


Téléchargements par mois depuis la dernière année

Plus de statistiques...

Benmouna, A., Becherif, M., Boulon, L., Dépature, C. et Ramadan, H. S. (2021). Efficient experimental energy management operating for FC/battery/SC vehicles via hybrid Artificial Neural Networks-Passivity Based Control. Renewable Energy, 178 . pp. 1291-1302. ISSN 0960-1481 DOI 10.1016/j.renene.2021.06.038

[thumbnail of BOULON_L_218_POST.pdf]
Télécharger (1MB) | Prévisualisation


Nowadays, the energy management of multisource hybrid systems is becoming an interesting and challenging topic for many researchers. The judicious choice of the energy management strategy not only allows for the best distribution of energy between the different sources, but also reduces the system's consumption, increases the life span of the used sources and fulfills the energy demand that affects the autonomy of the electric vehicle (EV). A novel hybrid control strategy based on the interconnection and damping assignment passivity-based control (IDA-PBC) technique is proposed while considering the battery State of Charge (SOC) and the hydrogen level operating conditions. PBC is a very powerful nonlinear technique, which uses important system information such as the system energy information. The Artificial Neural Network (ANN) is used for defining the appropriate references for the proposed controller to properly share the load power demand among the sources. Consequently, the proposed nonlinear control enables dispatching the requested power/energy among sources under source limitations. The real time experimental results demonstrate the enhanced efficiency of the hybridized ANN together with the IDA-PBC control. This work proposes a complete study and solution, from modeling, control, stability proof, simulation to practical validation. New constraints are emerging in anticipation of the real-time use of FC hybrid systems. These constraints and objectives are mainly related to the limitations of energy resources and the minimization of hydrogen consumption. The supervision of hydrogen level and battery SOC resources are proposed by using ANN, which gives the battery current and/or SC set point to the control loops. Experimentation works have validated the feasibility of this optimization technique.

Type de document: Article
Mots-clés libres: energy management hybrid system passivity based control artificial neural network
Date de dépôt: 24 août 2021 18:01
Dernière modification: 28 juin 2023 04:00
Version du document déposé: Post-print (version corrigée et acceptée)
URI: https://depot-e.uqtr.ca/id/eprint/9710

Actions (administrateurs uniquement)

Éditer la notice Éditer la notice