Search

International audience; In this work, the dynamic responses of the on board hydrogen storage system with commercially available metal hydride tanks are investigated based on the database collected from fuel cell electric vehicles operation experiments. A mathematical model considering the heat transfer measured by the temperature control system is developed to analyze the absorption and desorption reaction during operation. As a seal container filled with unknown metal hydride, the practical used hydrogen storage tank is a gray box in the embedded storage system. Without the information about characteristics of material, current operation state and degradation degree, the parameters of the model are uncertain. Particle swarm optimization (PSO) algorithm is applied to search for the optimal parameter set. The simulation results of the hydrogen storage system model using these parameters show great agreements with the real operating data under different temperature conditions; the maximal error is lower than 9%.

International audience; This paper proposes a novel Hermite neural network-based second-order sliding-mode (HNN-SOSM) control strategy for the synchronous reluctance motor (SynRM) drive system. The proposed HNN-SOSM control strategy is a nonlinear vector control strategy consisting of the speed control loop and the current control loop. The speed control loop adopts a composite speed controller, which is composed of three components: 1) a standard super-twisting algorithm-based SOSM (STA-SOSM) controller for achieving the rotor angular speed tracking control, 2) a HNN-based disturbance estimator (HNN-DE) for compensating the lumped disturbance, which is composed of external disturbances and parametric uncertainties, and 3) an error compensator for compensating the approximation error of the HNN-DE. The learning laws for the HNN-DE and the error compensator are derived by the Lyapunov synthesis approach. In the current control loop, considering the magnetic saturation effect, two composite current controllers, each of which comprises two standard STA-SOSM controllers, are designed to make direct and quadrature axes stator current components in the rotor reference frame track their references, respectively. Comparative hardware-in-the-loop (HIL) tests between the proposed HNN-SOSM control strategy and the conventional STA-SOSM control strategy for the SynRM drive system are performed. The results of the HIL tests validate the feasibility and the superiority of the proposed HNN-SOSM control strategy.

International audience; In this paper, an appropriate control technique is investigated to reduce vibration and noise in switched reluctance motors (SRM). Vibrations are mainly caused by radial force acting on salient poles of the motor. In order to determine this force, a no linear model for switched reluctance motor is developed, based on finite element calculation of inductance and natural frequencies of stator. Thus, the force component at a frequency close to natural frequency of stator should be reduced. In this context, a method of two steps commutation is also applied. To verify the effectiveness of the proposed method, the electro-mechanical-acoustic developed model makes possible to define theoretical spectrum of vibration and noise with conventional and proposed method. The simulation results showed that the proposed method performs better in terms of vibration and noise reduction.

With the further deterioration of environment and the depletion of fossil fuels, the alternative energy sources are urgently needed to be discovered. Hydrogen holds great promise thanks to its unlimited resources, high energy density and the environmentally friendly nature. However, its low volume density under normal temperature and pressure becomes the main challenge for on-board storage. Owing to its high potential of safety, one of the optimal solution for the future hydrogen vehicle is storing hydrogen using metal hydride (MH) under proper temperature and pressure. This work focuses on the state of charge (SOC) estimation of the embedded MH hydrogen storage tank. High precise estimating of the remaining energy will contribute to both the evaluation of reliability and the design of control strategy. A statistical model of SOC is proposed based on the database collected form laboratory experiments and real operation vehicle test. What’s more, a joint multi-classifier is designed to recognize the current state of reaction. Under this condition, the SOC of MH hydrogen storage tank is calculated through combining the state classifier and SOC model. This proposed on-line SOC estimation procedure is validated with the real operation vehicles in both charging and discharging process. It is proved to be effective with the estimation error of 0.2% during charging and 4.29% of discharging.

This paper proposes a finite-control-set model predictive current and capacitor voltage control strategy for the synchronous reluctance motor drive systems fed by the eight-switch three-phase inverter (ESTPI), which is a type of post-fault reconfigured three-level neutral-point-clamped inverter with open-circuit fault in a leg. The proposed control strategy uses the discrete-time models of the drive system to predict the behavior of the stator currents and the neutral point potential (NPP) error for zero and small voltage vectors provided by the ESTPI. The errors between stator currents and their references in the rotor reference frame as well as the NPP error are evaluated in a predefined cost function. The basic voltage vector, which minimizes this cost function, is selected. The proposed control strategy can achieve not only the reference stator current tracking but also the capacitor voltage offset suppression. Simulation results have shown the effectiveness of the proposed control scheme.

International audience; This paper proposes a rotor speed and position estimation scheme for the synchronous reluctance motor (SynRM) drive system based on its active flux model in the stator reference frame and the total least squares (TLS) EXIN neuron. Firstly, the active flux model of the SynRM in the stator reference frame is reconstructed to the overdetermined matrix equations. On the basis of that, the estimation of the rotor speed of the SynRM is transferred into solving a TLS problem. The TLS EXIN neuron, which is a recursive TLS algorithm, is used to solve this problem online to extract the rotor speed. The estimated rotor position is obtained from the estimated rotor speed based on the integrator. The feasibility and effectiveness of the proposed rotor speed estimation scheme have been verified by the simulation results.

International audience; A fuel cell power system integrating proton-exchange-membrane fuel cell (PEMFC) and metal hydride (MH)-based hydrogen storage tank presents great potential in transportation applications. The embedded PEMFC and MH tank are thermally coupled through a heat ex-changer and control system. The hydrogen generating and supplying rate from the MH tank to PEMFC is strongly influenced by the transferred heat, which affects the performance of long-term operation as well. In this work, the dynamic behavior of the fuel cell system is simulated with a mathematical model set and validated using a database from the real operation vehicles. Thanks to the heat ex-changer combined by fan, radiator and circulation water, the hydrogen flow rate from MH tank to PEMFC is well controlled to meet the requirement of power load. The simulated model describes the response of each component including the power and heat generated by PEMFC, the hydrogen desorption kinetics and the heat transfer in the system. A thermal management strategy with a PID controller is proposed to reduce the degradation and extend the lifespan of PEMFC. The results demonstrate that an optimized performance of PEMFC after 1000 h is realized. In spite of the MH tank degradation rate has been raised 0.3%, 2.5% of voltage degradation of PEMFC is reduced. Meanwhile, for the integration fuel cell system, more than 3% of fuel efficiency and 10% of fuel economy is saved.

International audience; The aim of this paper is to present an on-line adaptive equivalent consumption minimum strategy (AECMS) for fuel cell hybrid electric vehicle powered by fuel cell, battery and supercapacitor. In order to design the AECMS, an equivalent consumption minimum strategy (ECMS) without considering power sources degradation is firstly designed to decrease hydrogen consumption and degradation of power sources, which chooses fuel cell as the main power source to supply steady power, battery as the main energy storage source to buffer energy demand by vehicle and supercapacitor as the peak power supplier. A testbench is built to validate the developed ECMS. By contrastive experimental tests on ECMS, Rule Based Control Strategy (RBCS) and a Hybrid ECMS Operating mode control Strategy (HEOS) through the built test bench, hydrogen consumption of ECMS decreases 2.16% and 1.47% respectively and it also has the smoothest fuel cell current. Along with the degradation of fuel cell and battery, the charge sustenance objective of battery cannot be reached. Therefore, AECMS is finally designed to adjust equivalent factors and fuel cell dynamic current change rate along with the state of health (SOH) of fuel cell and battery, to make sure the charge sustenance of battery and prolong the lifetime of fuel cell. The method that on-line estimates their SOHs and the effects of their degradation on ECMS are also analyzed.

International audience

AC-DC converter switches of the drive train of series hybrid electric vehicles (SHEVs) are generally exposed to the possibility of outbreak open-phase faults because of troubles with the switching devices. In this framework, the present study proposes an artificial neural network (ANN)-based method for fault diagnosis after extraction of a new pattern. The new pattern under AC-DC converter failure in view of SHEV application has been used for train-proposed ANN. To achieve this goal, four different levels of switches fault are considered on the basis of both simulation and experimental results. Ensuring the accuracy and generalisation of the introduced pattern, several parameters have been considered, namely: capacitor size changes, load, and speed variations. The experimental results validate the simulation results thoroughly.

The aim of this paper is to present a sequential quadratic programming (SQP) based equivalent consumption minimum strategy (ECMS) (SECMS) for fuel cell hybrid electric vehicle (FCHEV) powered by fuel cell, battery and supercapacitor. In order to decrease hydrogen consumption and increase the durability of power sources, fuel cell is chosen as the main power source and supplies steady current, battery is designed as the main energy buffer and the replacement of fuel cell failure and supercapacitor is operated to supply peak power. Low energy density of supercpacitor lets its equivalent hydrogen consumption be taken as zero for many ECMS researches. This simplification leads to suboptimal fuel economy and complex of control system. SECMS considers hydrogen consumption of three power sources into objective function to solve this problem. A rule based control strategy (RBCS) and an hybrid ECMS operating mode control strategy (OMCS) (HEOS) are also designed to compare with SECMS. An experimental test bench is built to validate the comparative study of three strategies. The results show that compared with RBCS and HEOS, hydrogen consumption of SECMS decreases of 2.16 % and 1,47 % respectively and it also has the most smooth fuel cell current, which means a lowest fuel cell degradation.

International audience; In stand-alone power supply including a fuel cell (FC) and photovoltaic panels, DC/DC converters must meet several requirements from compactness, high voltage ratio, input current ripples reduction, high energy efficiency and reliability regarding keeping its function in case of power switch failures point of view. In this paper, a floating interleaved DC/DC boost converter has been chosen in order to meet the requirements for FC and photovoltaic applications. Despite the presence of a natural redundancy, a power switch failure may drastically reduce the reliability of the converter due to the additional stress on the electrical components. The purpose of this paper is to propose a power switch fault detection method based on Park's vector and a management of operating degraded mode. Obtained simulation and experimental results allow demonstrating the performance of the developed fault detection method to enhance the overall reliability of the converter.

International audience; Energy saving and environmental protection are two main themes for automobile development in the world. Fuel cell hybrid electric vehicle (FCHEV) aims to achieve these two goals. Since multiple power sources including fuel cell and energy storage sources are used to power the vehicle, an energy management strategy (EMS) plays an important role to improve the vehicles performance. Objectives should be pre-set in the process of designing EMS. According to the analysis of characteristics of FCHEV, classification of objectives and methods to achieve these objectives for different control strategies are presented and concluded.

International audience; Metal hydrides have been receiving a great attention as an effective and safe way to store hydrogen. Metal hydride-based hydrogen tank has a huge potential for the application of on-board hydrogen storage on the fuel cell electric vehicles. When the metal hydrides aging with cycle, the capacity of the hydrogen tank decreases. In this work, a mathematical model to estimate the hydrogen supply capability and observe degeneration conditions with cycle life of a hydride tank is presented. The used method is based on the exploitation of properties of Pressure-Composition-Temperature (PCT) isotherm of hydrogenation process in the hydride tank and the empirical modeling using experimental data. The established model can be used to describe the hydrogen supply capability and the concentration in the tank with equilibrium pressure and temperature under static state. This model could also be used to determine the aging degree of the tank after several charge and discharge cycles.

Stator turn faults in permanent magnet synchronous motors (PMSMs) are more dangerous than those in induction motors (IMs) because of the presence of spinning rotor magnets that can be turned off at will. Condition monitoring and fault detection and diagnosis of the PMSM have been receiving a growing amount of attention among scientists and engineers in the past few years. The aim of this study is to propose a new detection technique of stator winding faults in a three-phase PMSM. This technique is based on the image analysis and recognition of the stator current Concordia patterns, and will allow the identification of turn faults in the stator winding as well as its correspondent fault index severity. A test bench of a vector controlled PMSM motor behaviors under short circuited turn in two phases stator windings has been built. Some experimental results of the phase to phase short circuits have been performed for diagnosis purpose.

International audience; In recent years, the impact of the current ripple on Fuel Cell (FC) performances, particularly on its lifetime has gained a growing interest from the international scientific community. Polymer Electrolyte Membrane Fuel Cells (PEMFC) usually require a DC/DC boost converter to increase the output FC voltage. The output FC current is then submitted to the high switching frequency (i.e. &gt10 kHz) leading to a current ripple. Generally, in order to study the impact of the current ripple on FC performances, current ripple coming from DC/DC converters is simulated by means of an electronic load and signal generator. However, the use of a signal generator limits the scope of investigation up to a frequency of the current ripple equal to 10 kHz and by considering a healthy operation mode of the converter. For this reason, several aspects of the interaction between FC and DC/DC converter (operating at more than 10 kHz) require further investigation for automotive applications where the switching frequency is increased in order to reduce the volume of the converter, particularly in case of a faulty operation mode when power switch fault occurs from the DC/DC converter. The purpose of this paper is to provide simulation and experimental results at a switching frequency equal to 20 kHz, regarding the impact of open circuit fault of a power switch on the FC behavior, efficiency, H2 consumption, reliability and performance.

International audience; The accurate estimation of state of health (SOH) of Lithium-ion battery is critical to guarantee safety and reliability of whole system. Large number of researches on SOH have many strict constrained assumptions, such as repeated same load, which differ from real situation. In this paper, the study focus on variable load profile. An artificial neural network is designed to determine the battery SOH using operation times for 12 different currents and the beginning capacity for every cycle. Two kinds of validation samples are chosen to test the neural network. Simulation results show that for randomly chosen 20 validation samples from whole dataset, each sample errors don't exceed 4.37% and root mean squared error is 1.33%, for 20 sequential validation samples from one battery, maximum estimation error is 4.96% and root mean squared error is 1.25%, all below 5%, proving the good behavior of the designed neural network SOH estimation.

International audience; The objective of this paper is to design an equivalent consumption minimization strategy (ECMS) for a fuel cell hybrid electric vehicle powered by fuel cell, battery and supercapacitor. Fuel cell efficiency penalty coefficient, battery SOC coefficient and supercapacitor SOC coefficient are designed for ECMS to operate the fuel cell system at its best efficiency zone, the battery as a long term energy buffer and supercapacitor supplying peak power. This paper also takes fuel cell degradation into consideration and sets adaptive fuel cell current change rate to ensure the proper work of the fuel cell during all its lifetime and normal operation of vehicle under fuel cell degradation state.

International audience; The aims of this work is to appreciate the degree of relevance of the use the external magnetic field for diagnosis applications in the field of electrical traction provided by external rotor permanent magnets synchronous machines. A numerical modeling using the finite element method has been developed to simulate the behavior of the machine in operating generator in healthy and in faulty modes. Furthermore, the principle of the method of measuring the external magnetic field and the optimization of the magnetic sensors positions are explained. The study was validated by an experimental approach.

International audience; This paper describes a novel equivalent consumption minimization strategy (ECMS) for a fuel cell hybrid electric vehicle. The vehicle is composed of a fuel cell as main energy source, batteries and supercapacitors as peaking power sources. Through using fuel cell efficiency penalty coefficient, battery SOC coefficient and supercapacitor SOC coefficient, the designed new strategy aims to operate the fuel cell system at its best efficiency point, the battery as a long term energy buffer and supercapacitor to supply peak power. Considering the characteristic differences between battery and supercapacitor, different SOC penality coefficients are set to full play the potential of battery's large storage ability and supercapacitor's fast charge/discharge power. The proposed energy management is validated by simulation using driving cycle with urban pattern.

International audience; Due to the low and unregulated voltage generated by fuel cells, power electronic conditioning systems such as DC/DC converter are required in Fuel Cell Systems (FCS). These FCS are used in different applications with critical loads such as automotive applications, material handling equipment, and backup power. In these applications, it is crucial to guarantee reliability and availability. In FCS, DC/DC converters are one of the most important failure sources. Since the power switches ranked the most delicate components in DC/DC converters, the development of power switch fault detection algorithm is a mandatory step in order to ensure the availability of the system. The purpose of this paper is to propose a fast and original power switch fault detection algorithm based on Park's vectors combined with a Fault-Tolerant Control (FTC) for a 3-leg Interleaved DC/DC Boost Converter (IBC) used in FCS. The developed power switch fault detection and FTC are implemented on a FPGA target, allowing detecting, identifying and handling quickly the faulty power switch. The obtained results from experimental tests confirm the excellent performances of the proposed power switch fault detection algorithm and the FTC.

International audience; In recent years, the effects of high-frequency current ripple coming from DC/DC converters on Proton Exchange Membrane Fuel Cell (PEMFC) have gained a growing interest from the international scientific community. Currently, the durability is one of the major technical challenges to PEMFC commercialization. Basing on the existing literature, the high-frequency current ripple leads up to long-term degradations on PEMFC, particularly on its lifetime. Accordingly, in order to enhance the PEMFC lifetime, new DC/DC converter topologies have been developed, such as Interleaved DC/DC Boost Converter (IBC). Despite this topology can minimize drastically the current ripple, the loss of one leg in case of power switch fault leads up to the drastic increasing of the current ripple and consequently long-term degradations on the PEMFC. In order to cope with this issue, solutions have to be developed. Within the framework of this research work, an efficient Fault-Tolerant Control (FTC) implemented on an FPGA board has been developed to solve this issue. The latter consists in changing the PWM gate control signal according to the faulty leg given by a fault detection algorithm. The obtained experimental results between a PEMFC and an IBC topology allows demonstrating the ability of the FTC to reduce drastically the current ripple in case of power switch faults.

International audience; Over the last decade, fault-tolerance has gained a growing interest from the scientific community, particularly for Proton Exchange Membrane Fuel Cell (PEMFC) and its DC/DC converter. Fault-tolerance has two main objectives: provide highlevel improvement in system availability and lead to substantial reduction in maintenance costs. Generally, fault-tolerance refers to a Fault-Tolerant System (FTS) and/or Control (FTC). PEMFCand its DC/DC converter must respond to challenging issues for stationary and automotive applications, in terms of reliability. Indeed, the unavailability of the current Fuel Cell System (FCS) stops to market these components on a large scale. In PEMFCs, the Membrane Electrode Assembly (MEA) can be subjected to different failures: membrane break, internal gas leakage, cell flooding or drying, poisoning of the catalyst areas… By comparison, in DC/DC converters, power switches ranked the most delicate components. Once the failure has been detected,FCS stop must be avoided. In order to ensure the availability of the FCS and/or to minimize the undesirable effects of degraded operating mode on FCS, FTS and FTC must be used. This paper presents a FTC applied to the DC/DC converter in case of drying membrane failure in PEMFC. The FTC consists in degrading the converter performances through its control in order to bring back the PEMFC in an optimum operating condition. Thedeveloped FTC is implemented experimentally on a MicroAutoBox. The obtained experimental results demonstrate the efficiency of the FTC to improve the performances of the PEMFC.

Permanent magnet synchronous machines (PMSM) have high efficiency and torque density, and have already been employed in hybrid electric vehicles. However, one of their disadvantages is the inherent cogging torque, which is a kind of torque ripple and it would be better to minimize. This torque, sometimes, can be an important source of noise and vibrations. In this paper, the effect of the geometric characteristics of the stator on the vibratory behavior of electrical machines is illustrated. The optimum geometry for obtaining a minimum vibration level has been reached. For this purpose, an approach by using the Artificial Intelligent (AI) and the Finite Element Method (FEM) is proposed to solve the magneto-mechanical problem of geometrical parameters identification in the optimization process. The obtained results by Genetic Algorithm (GA) method have been presented.

Due to low DC voltage generated from fuel cell, a DC/DC converter must be used to interface fuel cell and DC bus in automotive applications such as fuel cell electric vehicles (FCEVs). Furthermore, the occurrence of power semiconductors failures in DC/DC converter has negative effects both for fuel cell and DC/DC converter and consequently on the whole powertrain. The purpose of this paper is to show and analyze the impacts of power semiconductors faults on energy efficiency and waveforms (e.g. FC and DC bus voltages, FC current) without modifying the topology and control. Besides, it has been chosen to focus this study on two topologies of interleaved DC/DC converters enabling to take into consideration the requirements of FCEV applications. These topologies have the capacity compared with the boost topology to operate even in case of faults, ensuring consequently a continuity of service.

The paper presents a study concerning the influence of the capacitor value on the magnetic noise of the network and inverter-fed permanent split capacitor induction motor. To this aim, a numerical model of a 0.75 kW, 1500 rpm, permanent split-capacitor induction motor is developed under Flux2D Finite Element Software, taking into account the case of inverter supply by coupling Flux2D with Matlab/Simulink. Also, a mechanical analysis, using Ansys3D is performed on the studied motor. In order to verify the obtained results, noise measurements are performed on the network and inverter-fed single phase 0.75 kW, 1500 rpm, permanent split-capacitor induction motor fitted with different capacitor values (25uF, 36 uF, 50 uF).

This paper describes an electromechanical model for bearing fault modeling and detection in Permanent Magnet Synchronous Motors. Bearing problems are one major cause for drive failure. Their detection is possible by vibration monitoring of characteristic bearing frequencies. In the new approaches proposed in this work, the vibrations of the rotor shaft due to bearing failure are calculated by solving Hertz's equation contact numerically. This vibrations leads to load torque variations. The spectral analysis of the Park's vector modulus shows the possibility of the model to detect bearing faults.

The paper presents a comparison between the obtained results from testing a special design 5 kW squirrel-cage three-phase asynchronous machine and the numerical model of the same machine, simulated under the FLUX2D Finite Element Software. The aim is to validate the numerical model in the loaded and unloaded cases. The resulting model will be used for further research into the vibratory behaviour and noise of the motor. In this paper, an experimental approach is used to show the influence of load on motor vibration and noise levels. The relation between the natural frequencies of the stator, the frequencies of the forces causing vibration and the noise of the motor, is studied by analysing several sets of measurements.

The paper deals with the characterization of a permanent magnet synchronous machine (PMSM) used in electric traction applications. First of all, the main data (the output power and desired speed) of the electric traction application are defined. Starting from these elements, the PMSM is characterized by means of analytical, numerical and test obtained results. Thus, several design elements are stated in order to characterize globally the studied PMSM. Next, the control for such a motor is employed by means of numerical computation (by using the finite element method and the field oriented control technique). Finally, the PMSM for electric traction application is tested on bench, the obtained results being satisfactory.

This paper deals with the active compensation of the torque harmonics in brushless DC motors. The proposed method consists in analytical computation of the currents whose allow to remove electromagnetic torque harmonics. A first interest of this paper is that, for a given motor, the number of currents harmonics is minimized. Therefore, the maximum frequency of currents is limited, so that practical implementation is easier, A second and important interest is that harmonics multiple of three are not considered for the current computation so that the motor phases can be star-connected and the motor can be fed with a classical three-phase full-bridge inverter and not with three single-phase inverters. The complete methodology of the currents harmonics computation is presented. To illustrate the method, an example of calculation concerning a motor integrated in the wheel of a bicycle is detailed. Then, using a special test bench, the real machine is fed with the optimal currents in order to confirm the possibility of experimental implementation. Finally, experimental measurements concerning the vibratory behavior are proposed and the results obtained with the optimal fed are compared with the ones obtained with a 120/spl deg/ square fed.

The paper proposes a method for modeling electromagnetic forces and analyzing the vibrations in synchronous motors with permanent magnets. A theoretical numerical model (based on finite element analysis of magnetic field and MATLAB/SIMULINK uses) was developed permitting study of the rotor's vibration and the effects of the motor feed on it. We propose the realization of an algorithm that allows the calculation of the current to obtain a better behavior of the motor. It is considered that the radial forces give the vibratory behavior and the ripples of the tangential force produce ripples of the torque.

This paper deals with active compensation of torque harmonics in brushless DC motors. The proposed method consists in analytical computation of currents which allow removal of electromagnetic torque harmonics. The originality of this paper is that, for a given motor, the number of harmonic currents is minimized. Therefore, the maximum frequency of currents is limited so that practical implementation is simplified. A complete algorithm of currents harmonics computation is presented and an example concerning a motor integrated in the wheel of a bicycle is detailed, in order to show the accuracy of the method.

Nowadays, the development of alternative energy sources becoming particularly important due to the effect of climate change and fossil fuels depletion. Hydrogen holds great promise thanks to its unlimited resources, high energy density, large technological flexibility and the environmentally friendly nature. With high potential of safety, storing hydrogen with metal hydrides (MH) is considered to be the optimal on-board hydrogen storage method for the future hydrogen vehicle. This thesis therefore contributes to analyzing the performance and proprieties of embedded MH hydrogen storage systems, including the characteristic estimation, dynamic modeling and thermal management coupling with fuel cells.Firstly, statistical models are proposed for dynamic performance analysis and state of charge (SOC) calculation. The online SOC estimation process is then realized combining a multi-joint state classifier. The dynamic model of the embedded MH tank considering mass and energy conversion is proposed using optimized parameters identified through particle swarm optimization (PSO) algorithm. Moreover, the dynamic behavior of the fuel cell system integrating proton-exchange-membrane fuel cell (PEMFC) and MH hydrogen storage tank is simulated with a mathematical model set and validated using a database from the real operation vehicles. A thermal management strategy with PID controller is proposed to reduce the degradation and extend the lifespan of PEMFC. Finally, a test bench is designed in laboratory and experiments are carried out to validate the proposed models and strategies., Le développement des sources d'énergie alternatives devient très important en raison de l'effet du changement climatique et de l'épuisement des combustibles fossiles. L’hydrogène est prometteur grâce à sa ressource illimité, sa densité énergétique élevée, sa grande flexibilité technologique et sa nature respectueuse de l’environnement. Avec un potentiel élevé en matière de sécurité et de fiabilité, le stockage d'hydrogène avec des hydrures métalliques (MH) est considéré comme la meilleure méthode. Cette thèse contribue à l'étude des performances des systèmes de stockage de l’hydrogène MH embouqué, plus particulièrement l’état de charge, la modélisation dynamique et la gestion thermique la pile à combustible.Tout d'abord, des modèles sont proposés pour l'analyse dynamique des performances et le calcul de l'état de charge (SOC). Le processus d'estimation SOC en ligne est ensuite réalisé en combinant un classifieur d'états multi-joint. Le modèle dynamique du réservoir prenant en compte la conversion de masse et d'énergie est proposé à l'aide de paramètres optimisés identifiés par un algorithme d'optimisation d'essaim partiel. De plus, le comportement dynamique du système de pile à combustible intégrant la pile à combustible à membrane échangeuse de protons (PEMFC) et le réservoir de stockage d'hydrogène MH est simulé à l'aide d'un modèle mathématique défini et validé à l'aide d'une base de données provenant des véhicules réels. Une stratégie de gestion thermique avec contrôleur PID est proposée pour réduire la dégradation et prolonger la durée de vie de la PEMFC. Enfin, d’essai est conçu en laboratoire et des expériences sont menées pour valider les modèles et stratégies proposés.