Control and Energy Management of an Electrical Vehicle

Abstract

ABSTRACT : This research focussed on two axes, the first one was the energy management of the electric vehicle’s power-train system , and the second one was the control of its drive-train. The first part of study was to address the problem of the control of DC-bus voltage including fuel cell and supercapacitor energy sources. We started by developping a control strategy of a fuel cell stack using a non-isolated DC/DC converter. This control was ensured by hybrid duel loop control, which included a voltage loops with a inear PI controllers and a fast current loops with a non-linear sliding controllers. The mplementation of the developped FC’s control system was made using an experimental est bench of the studied converter, which was carried out on an experimental bench at reduced power (120 watts) of equipment available in our laboratory MSE. Then, an energy management method based on the flatness strategy and the sliding mode control were suggested and explained to control a hybrid system composed of a fuel cell and a upercapacitor as the main and auxiliary sources. The nonlinear flatness strategy was used to achieve a linearising feedback control law that gives an exponential tracking of he FC and SCs power trajectories. The proposed flatness control allows decoupling the hybrid system into two decoupled sources so as each subsystem has a separated control larget expressed in terms of a sliding surface. The role of sliding mode controllers was o ensure the power sharing between the DC-link inverters by controlling the FC and SC currents. The control of the drive-train was to provide an inversion based control and propose an adaptive operation mode of the studied electric vehicle (EV) by using a fuzzy ogic to invert the accumulation element. This model-based control was realized by means of formalism called Energy Macroscopic Representation (EMR). Hence , EMR gives insights into the real energy operation of the EV system and allows a deep understanding of its potentials from dynamics point. Also, the EV has been modelled by a linear dynamic model expressed by a state space representation in order to control its velocity by a states feedback controller. The control systems were simulated in Matlab/Simulink and several simulation tests were carried out for the European, the American driving cycles. They were made in order to validate the functionality of the designed control systems.