| KEYWORD |
MPC control of a FCEV in a real-world traffic scenario
Parole chiave FUEL CELL, OPTIMIZATION ALGORITHMS, TRAFFIC, VEHICLE DYNAMICS
Riferimenti HENRIQUE DE CARVALHO PINHEIRO
Gruppi di ricerca 27- Veicoli innovativi elettrici e ibridi
Tipo tesi ALGORITHMS, OPTIMIZATION, SIMULATION
Descrizione The automotive industry is in a phase of profound renewal as the large-scale adoption of V2X technologies to improve energy efficiency of vehicles is being investigated. Indeed, the availability of traffic information allows to optimize vehicle speed profile and power management based on real-time driving conditions.
Assessing the impact of such technologies on vehicle energy efficiency is challenging as statutory driving cycles, traditionally used for vehicle homologation purposes, are not suitable for the scope. To solve this problem, a traffic-in-the-loop simulation environment has been developed, capable of simulating vehicle performance in real-world scenarios. The selected candidate will rely on such pre-existing tool and on a FCEV model developed in Simulink for the development of the project. The primary scope of the thesis is to build an MPC controller based on Dynamic Programing for splitting the power request between the battery pack and the fuel cell stack.
After an initial phase of literature review about the state of the art of traffic-in-the-loop vehicle simulations and FCEV energy management strategies, the selected candidate will start working on the implementation of the controller within the traffic-in-the-loop simulation framework using a pre-existing traffic scenario. The project will conclude by analyzing how vehicle performance changes with variable conditions, including the effect of factors like traffic congestion and traffic lights schedules.
The thesis work will require at least the following tasks:
· Development of a Matlab function containing the MPC controller
· Implementation and testing of the MPC in the Co-Simulation environment
· Analysis of the results
If the work progress is fat, these further goals might be included:
· Definition of a simulation plan with the MPC to generate data for training the AI controller
· AI controller training
· Definition of a simulation plan with the AI controller to collect results
· Implementation of the AI controller in place of the MPC
· Comparison between the outcomes
Conoscenze richieste - Software: knowledge of MATLAB/Simulink is mandatory.
- Language: knowledge of English is mandatory.
- Personal skill: proactive attitude to problem solving.
Scadenza validita proposta 28/11/2025
PROPONI LA TUA CANDIDATURA