• Energy Research

Le véhicule hybride est un système complexe constitué de plusieurs sources et dont l'agencement de sa transmission peut faire appel à plusieurs composants. La possibilité d'avoir des dimensionnements différents de la batterie par rapport au moteur thermique, couplée aux diverses topologies possibles, représentent autant de degrés de liberté qui peuvent être exploités pour son optimisation énergétique. Le premier chapitre de ce mémoire expose cette diversité en présentant les différentes classifications usuelles des Îhicules hybrides. Mises à part quelques applications qui visent à augmenter les performances dynamiques des Îhicules (vitesse et accélération), l'objectif de l'hybridation est principalement la réduction de la consommation énergétique et des émissions de polluants. C'est l'objectif énergétique qu'on considère dans ces travaux. Afin d'atteindre cet objectif, des optimisations sont nécessaires sur différents plans. Pour un usage donné, les performances énergétiques du Îhicule hybride dépendent de trois aspects fortement interdépendants qui sont i) la topologie (série, parallèle, mixte) ii) le dimensionnement des composants iii) la stratégie de gestion de l'énergie entre les différentes sources. Pour appréhender ces différentes dimensions et leur couplage, une approche systémique s'appuyant sur la modélisation a été mise en place avec les membres de l'équipe Véhicules électriques et hybrides du LTE. Cette approche a abouti au développement d'un outil de simulation, VEHLIB, qui a permis de capitaliser les différents travaux de modélisation. Cet outil, présenté dans le deuxième chapitre, a été utilisé ensuite pour servir les objectifs d'optimisation de la gestion de l'énergie et d'aide au dimensionnement optimal des Îhicules hybrides. Dans le cadre du réseau MEGEVH , une ouverture vers la Représentation Energétique Macroscopique (REM, développée au L2EP) a permis de démontrer l'apport incontestable de la REM pour la synthèse systématique de la commande des systèmes. Le chapitre III, consacré à la gestion de l'énergie, présente un état de l'art des méthodes développées ces dix dernières années, ainsi que notre contribution dans ce domaine. Cette dernière a consisté dans un premier temps en l'utilisation et l'amélioration des méthodes à base de règles expertes. Ensuite deux thèses ont proposé l'optimisation de la gestion de l'énergie de point de vue de la consommation de carburant. Tous ces travaux se sont appuyés sur une démarche utilisant la modélisation, ainsi que l'expérimentation sur banc d'essai à l'échelle 1 dans une configuration de Îhicule émulé (Hardware In the Loop - HIL - simulation). Il s'est posé également, pour le Îhicule hybride, le problème de la taille relative de la batterie et des machines électriques vis-à-vis de la taille du moteur thermique. En effet, pour un cahier des charges dynamique donné, plusieurs dimensionnements peuvent être admissibles. Une procédure d'aide au dimensionnement optimal a été mise en oeuvre dans l'équipe et a fait l'objet de travaux exposés dans le chapitre 4. En théorie, la définition générale du Îhicule hybride ne se limite pas à la version thermique - électrique qui a fait l'objet de la plus part de nos contributions jusqu'ici. D'autres possibilités d'association de sources (Pile à combustible, supercondensateurs, volent d'inertie, ...) sont envisagées et font l'objet de travaux récents. On peut parler dans ce cas de Îhicule multi-sources. Tant au niveau des topologies, qu'au niveau de la gestion de l'énergie et du dimensionnement, des travaux de recherches sont encore nécessaires pour tenter de généraliser les concepts déjà développés pour le Îhicule hybride thermique-électrique. L'utilisation de formalismes structurants à l'image de la REM permettrait d'appréhender la complexité croissante et de réaliser l'articulation entre les différents niveaux de commande, locale et globale. Sur ces différentes dimensions, des ouvertures et des perspectives sont détaillées dans le dernier chapitre de ce mémoire. The Hybrid Electric Vehicle is a complex system made up of several sources and the arrangement of its transmission can appeal to several components. The possibility of having different sizes of the battery compared to the combustion engine, coupled with the various possible topologies, represent as many degrees of freedom that can be exploited for its energy optimization. The first chapter of this paper exposes this diversity by presenting the different common classifications of hybrid electric vehicles. Apart from a few applications that are intended to increase the dynamic performance of the vehicle (speed and acceleration), the goal of hybridization is mainly the reduction of energy consumption and emissions of pollutants. It is the energy target that is considered in this work. To achieve this goal, optimizations are needed on different plans. For a given use, the energy performance of the hybrid vehicle depends on three strongly interdependent aspects which are i) topology (series, parallel, dual) ii) the sizing of components iii) energy management strategy to share the instantaneous power demand. To understand these different dimensions and their coupling, a systemic approach based on modelling was implemented with the Electric and Hybrid Vehicle team members. This approach has led to the development of a simulation tool, VEHLIB, that allowed to capitalizing different modelling works. This tool, which is presented in the second chapter, has then been used to serve the objectives of the energy management optimization and support the optimal design of hybrid vehicles. Under the MEGEVH network[1] , an opening towards the Energetic Macroscopic Representation (EMR, developed by the L2EP) showed the undeniable contribution of the EMR to the systematic synthesis of complex systems' control. Chapter III, dedicated to energy management, presents a State of the art of the methods developed these past ten years, and our contribution in this area. The latter was initially in the use and improvement of the rule based methods. Then two theses under my supervision proposed the optimization of the energy management in terms of fuel consumption. All these works relied on an approach using modeling, as well as experimentation on a test bench in an emulated vehicle configuration (Hardware In the Loop - HIL - simulation). Was also highlighted for the hybrid electric vehicles the problem of the relative size of the battery and electrical machines compared to the size of the combustion engine. Indeed, for a given dynamic specifications, several sizing may qualify. A help to the optimal sizing procedure has been implemented in the team and has been the subject of work described in Chapter 4. Theoretically, the general definition of hybrid vehicles is not limited to thermal - electric version which was the subject of the majority of our contributions so far. Other possibilities for association of sources (fuel cell, supercapacitors, flywheel, ...) are being considered and are the subject of recent work. We can speak in this case of multi-source vehicle. Either at topology level or at the level of the energy management and components' sizing, research is still needed to try to generalize the concepts already developed for the hybrid electric vehicle. The use of structuring formalisms like the EMR would help to understand the growing complexity and achieve the articulation between the different levels of control, local and global. On these different dimensions, perspectives and opportunities are detailed in the last chapter of this report.

This paper deals with energy management for electric vehicle using SuperCapacitors (SCs) and battery. SCs are connected to the DC bus through a converter whereas battery is directly connected to the DC bus. Using the inversion-based rules of Energetic Macroscopic Representation (EMR), a systematic control structure can be deduced. Some differences can be shown in comparison with classical control schemes. This paper aims to compare a control scheme deduced from EMR and another control structure designed thanks to a more classical way. It can be noticed that some differences appear. As far as the structure is concerned, the number of sensor can be reduced in the EMR case. In terms of performances, some control errors appear in the classical scheme case.

Predicting the operating temperature of lithium-ion battery during different cycles is important when it comes to the safety and efficiency of electric vehicles. In this regard, it is vital to adopt a suitable modeling approach to analyze the thermal performance of a battery. In this paper, the temperature of lithium-ion NMC pouch battery has been investigated. A new formulation of lumped model based on the thermal resistance network is proposed. Unlike previous models that treated the battery as a single entity, the proposed model introduces a more detailed analysis by incorporating thermal interactions between individual cells and tabs within a single cell scenario, while also considering interactions between cells and insulators or gaps, located between the cells, within the module case. This enhancement allows for the precise prediction of temperature variations across different cells implemented within the battery module. In order to evaluate the accuracy of the prediction, a three-dimensional finite element model was adopted as a reference. The study was performed first on a single cell, then on modules composed of several cells connected in series, during different operating conditions. A comprehensive comparison between both models was conducted. The analysis focused on two main aspects, the accuracy of temperature predictions and the computational time required. Notably, the developed lumped model showed a significant capability to estimate cell temperatures within the modules. The thermal results revealed close agreement with the values predicted by the finite element model, while needing significantly lower computational time. For instance, while the finite element model took almost 21 h to predict the battery temperature during consecutive charge/discharge cycles of a 10-cell module, the developed lumped model predicted the temperature within seconds, with a maximum difference of 0.42 °C.

Abstract The paper deals with real-time energy management strategies for a hybrid energy storage system including a battery and supercapacitor for an electric vehicle. Besides efficiency concerns, a key issue for real-time applications is ensuring safe operations for the considered system. In such a system it is mandatory to limit the supercapacitor voltage by the energy management strategy to comply with efficiency and safety constraints. An optimisation-based strategy (λ-control) and a rule-based strategy (filtering) are compared in this paper. The active limitation of the supercapacitor voltage is ensured in both strategies. Experimental results show that the two strategies have equivalent performances under real-driving cycles not known in advance. A difference of 2% between the strategies on the battery current root mean square value is shown. By varying the desired supercapacitor voltage range, the results show that the λ-control is better-suited for a high supercapacitor voltage range whereas the filtering gives the best performances for a low supercapacitor voltage range.

I. Hybrid Energy Storage System (HESS)II. Different topologies of HESSs batt. / supercap.III. Model of the studied HESSsIV. Control of the studied HESSsV. Example of results

Hybrid vehicles use two energy sources for their propelling. Usually an internal combustion engine (ICE) is used with one or more electric machine(s) (EM). The problem is then to split the driver power demand between the ICE and the EM in order to minimize a criterion, usually the fuel consumption. A global optimization algorithm based on optimal control theory is recalled. The obtained results are optimal but can only be obtained in simulation. For real time control purpose, this optimization algorithm is applied on a receding horizon. The main problem is then to choose the variables to be predicted on this horizon. By analyzing the optimization algorithm, it is shown that the prediction of the future driving conditions (vehicle speed and driver torque demand) is not necessary. Therefore, under some assumptions, a real time control is possible.

ADVANCED electric drive traction systems are being developed in order to ensure better energy efficiency for the emerging and future transportation systems such as electric vehicles (EVs), hybrid electric vehicles (HEVs), plug-in hybrid electric vehicles (PHEVs), fuel cell vehicles (FCVs), as well as electrified and advanced tractions and propulsions for trains, subways, ships, and airplanes. The research and development aims at reducing energy consumption and pollutant emissions in order to improve sustainability and address climate change. In this special section, The three first papers deal with new electric drives (a new electromagnetic active suspension, an electric variable transmission, and a charger-traction drive) where the design is realized by taking into account the constraints of the entire vehicle system. Two papers present the energy management of complex hybrid energy storage subsystems (fuel cell/supercapacitor and fuel-cell/battery/supercapacitor) using a new 'system description' [Energetic Macroscopic Representation (EMR)]. Two papers propose statistical studies of real-life drive cycles for the design and analysis of new low-carbon vehicles (plug-in hybrid electric vehicles and fuel cell vehicles).