Este trabalho insere-se na área de pesquisa de sistemas de transporte inteligente e mobilidade urbana buscando um cenário onde a infraestrutura rodoviária é capaz de monitorar um tráfego exclusivo de veículos inteligentes que não dependem de motoristas para serem guiados. A principal contribuição do trabalho é o desenvolvimento de uma solução matemática para otimizar o gerenciamento microscópico centralizado do tráfego de veículos inteligentes em trechos (segmentos) de rodovia. Para isto é apresentado um modelo de otimização baseado em Programação Linear Inteira Mista (MILP), que determina um plano ótimo de trajetórias individuais dos veículos em uma evolução de tráfego. O objetivo é reduzir o tempo de viagem individualmente e assegurar fluidez do tráfego. O modelo considera componentes essenciais do sistema dinâmico viário como topografia da pista, regras de trânsito e a curva de aceleração máxima de cada veículo. São contempladas várias situações de tráfego, tais como ultrapassagens, inclinação na pista, obstáculos e redutores de velocidade. Os resultados indicaram uma média de 20,5 segundos para o cálculo de um cenário com 6 veículos e 11 intervalos de tempo. Como o modelo MILP não tem solução em tempo computacional aceitável para aplicação real, também é proposto um algoritmo de simulação baseado em heurísticas o qual busca reduzir esse tempo de cálculo em detrimento da otimalidade da solução. O algoritmo reproduz o comportamento de um motorista que tenta manter sempre um valor de velocidade escolhido previamente, e por isso é forçado a ultrapassar outros veículos quando obstruído ao longo do trajeto. O resultado do algoritmo tem importância adicional, pois serve de referência para resolver o problema da prioridade nas ultrapassagens. Também são propostos novos indicadores para a avaliação microscópica de qualidade de tráfego. Finalmente, são apresentados resultados de testes em simulações a fim de avaliar e validar o modelo e o algoritmo. / This work focus on the research area of intelligent transportation systems and urban mobility. It considers a scenario where the roadside infrastructure is capable of monitoring traffic composed by 100% of intelligent vehicles that do not rely on drivers to be guided. The main contribution of this work is the development of a mathematical solution to optimize the centralized management of intelligent microscopic vehicular traffic in parts (segments) of highway. Therefore an optimization model based on Mixed Integer Linear Programming (MILP) is presented. The model determines individual trajectories plans of vehicles in a traffic evolution. The objective is to reduce the travel time individually and ensure traffic flow. The model considers essential components of the dynamic highway system, such as, topography of the lane, traffic rules and acceleration curve for each vehicle. Many traffic situations are considered, such as, overtaking, slopes, obstacles and speed reducers. The results indicated an average of 20.5 seconds to calculate a scenario with 6 vehicles and 11 time intervals. As the MILP model has no solution in acceptable computational time for real application, it is proposed an algorithm based on heuristic simulation which seeks to reduce the computation time at the expense of optimality of the solution. The algorithm reproduces the behavior of a driver who always tries to maintain a preselected velocity value, and is therefore forced to overtake other vehicles when blocked along the path. The result of the algorithm has additional importance because it serves as a reference for solving the problem of priority when overtaking. New indicators for microscopic evaluation of quality traffic are also proposed. Finally, test results are presented on simulations to evaluate and validate the model and algorithm.

A presente dissertação traz a questão da aderência de viagens de veículos em monitoramentos inteligentes com sistemas Automatic Vehicle Location (AVL) que operam em rotas rodoviárias. Uma viagem é considerada como uma série de \"passadas\", que correspondem ao tempo em que o veículo está em movimento, mas excluindo os tempos gastos em paradas para atividades como carregamento/descarregamento entre outras. A partir de dados históricos coletados via Global Positioning System (GPS) pelo sistema AVL, uma metodologia de filtragem e aplicações estatísticas para geração das passadas é apresentada. Além disso, são propostos métodos para geração de padrões de viagem de referência, baseados em tempos de viagem e velocidades, desvios padrões, locais de descontinuidades entre outros parâmetros. A geração desses padrões em conjunto com procedimentos operacionais permite o monitoramento eficiente do progresso de viagens de frotas de veículos, para finalidades logísticas e de segurança. O progresso de um veículo ao longo de uma rota é analisado diante dos padrões de viagem de referência obtidos a partir de suas viagens prévias, de veículos similares na mesma rota ou de viagens em rotas de mesma classe, dependendo do que for mais adequado. A geração de padrões é um processo dinâmico que gera conhecimento sobre o veículo e comportamento da rodovia ao longo do tempo. Desenho do processo de monitoramento do progresso de viagem é apresentado, no qual, a cada nova coleta de dado GPS ou a cada instante solicitado pelo usuário, a aderência é medida, eventuais descontinuidades (saídas da rota, paradas ou mudança de sentido) são identificadas e avisos são gerados. Tal aderência é definida por índice de desempenho que considera os desvios de tempo em relação a valores de referência e respectivas tolerâncias. Para experimentação da metodologia, foi realizada simulação de viagem na rodovia BR116 na ligação São Paulo - Rio de Janeiro, sobre base com cerca de 130.000 registros de dados GPS associados. Com integração em Geographic Information System (GIS) para suporte de funcionalidades, foram gerados os padrões de viagem e simulado o processo de monitoramento com sucesso. / This dissertation addresses the question of vehicle travel adherence in intelligent monitoring with Automatic Transportation Location (AVL) operating in a regional environment. A trip is considered as series of runs, corresponding to time in movement but excluding time spent on activities such as loading/unloading and others. Based on historic data collected from AVL/GPS a statistical data filtering method to generate the runs is presented. Furthermore, statistical methods are proposed to generate travel patterns based on travel time, speed, standard deviation and other parameters. The pattern generation together with operational procedures allows effective monitoring of large fleets in logistics and safety. The progress of a vehicle along a route is evaluated face to the statistical patterns of its previous successful trips or against statistical patterns of similar vehicles on the same route, whichever appropriate. The generation of patterns is a dynamic continuous process that generates knowledge on vehicle and road behavior along time. A broad outline of the travel monitoring process is presented. Whenever the requested by user, the process calculates the travel adherence, identifies abnormalities and generates alarms. That adherence is defined by a performance index, which considers the travel time deviations from the reference values and the respective tolerances. Successful experimentation was carried out on the Rio de Janeiro - São Paulo motorway, with 130.000 Global Positioning System (GPS) positional data relayed from trucks to a Geographic Information System (GIS) based monitoring system in Brazil.

Cette thèse est consacrée au domaine interdisciplinaire des Systèmes de Transport Intelligents et des technologies de Réalité Virtuelle. Elle se concentre sur l’amélioration des stratégies de commande des véhicules intelligents en tenant compte des impacts de l’environnement naturel ainsi que sur l’analyse de performance, la visualisation et la vérification de la validité des algorithmes de commande sur la plateforme de véhicules intelligents réalité virtuelle (IVVR).La plateforme IVVR comprend trois sous-systèmes : un sous-système de commande de véhicules intelligents, un sous-système de visualisation et un sous-système virtuel sans fil. Le synthétique environnement naturel a été modélisé et simulé pour la simulation et l’analyse de performance des stratégies de commande sous conditions environnementales complexes. Ensuite, les expérimentations concernant le trafic équipé du régulateur de vitesses adaptatives (ou coopérative) sont exécutés et ils montrent que les systèmes existants ont échoué à maintenir une espace inter-véhiculaire de sécurité lorsque les conditions d’environnement naturel sont défavorables. Dans ce cas, nous proposons un nouvel algorithme de commande appelé NECACC pour le contrôle longitudinal du véhicule en maintenant une espace inter-véhiculaire de sécurité et garantissant une capacité de circulation optimisée même dans des conditions environnementales complexes. Cet algorithme est ensuite simulé, vérifié et validé sur la plateforme IVVR. Enfin, les démonstrations de trafic virtuel effectuant des manœuvres communes de circulation contrôlés par les systèmes de commande intégrés proposées sont présentées sous diverse conditions environnementales / This thesis is dedicated to the interdisciplinary area of Intelligent Transportation Systems and Virtual Reality technologies. It focuses on the improvement of intelligent vehicles control strategies by considering the natural environment impacts as well as the visualization, the verification and performance analysis of proposed control algorithms on the proposed Intelligent Vehicles Virtual Reality (IVVR) platform.The IVVR platform includes three subsystems: Vehicle Intelligent Control Subsystem, Visualization Subsystem and Virtual Wireless Subsystem. For realizing the control strategy simulation and performance analysis under complex natural environment conditions, Synthetic Natural Environment has been modeled and simulated in this IVVR platform. Therefore, experiments of Adaptive Cruise Control (ACC) or Cooperative ACC system equipped traffic are executed and show that normal ACC/CACC system fails to keep a safe inter-vehicle space when the natural environment condition is variable or adverse especially in stiff conditions. For solving this problem, we propose a new control algorithm called NECACC (Natural Environment based CACC) for longitudinal vehicle control in maintaining a safe inter-vehicle distance as well as guaranteeing an appropriate traffic capacity even under complex environmental conditions. This algorithm is then simulated and verified in IVVR platform as a “proof of concept”. Finally, some virtual traffic demonstrations performing common traffic maneuvers are presented in IVVR platform under various environmental conditions. The vehicle platoon is controlled by proposed integrated control system and the safety can be ensured all the time

Les travaux présentés dans ce mémoire de thèse s’inscrivent dans le cadre des Systèmes de TransportIntelligents (STI). Bien que les premières études sur ces systèmes ont commencé dans les années 60, leurdéveloppement reposant sur les techniques de l’information et de la communication, a atteint sa maturitédans le début des années 80. Les STI, sont composés de différents systèmes et intègrent différents concepts(systèmes embarqués, capteurs intelligents, autoroutes intelligentes, . . .) afin d’optimiser le rendementdes infrastructures routières et répondre aux problèmes quotidiens des congestions. Ce mémoire présentequatre contributions dans le cadre du trafic routier et aborde les problèmes de l’estimation et de lacommande afin d’éliminer les problèmes de congestions « récurrentes ». Le premier point traite unproblème crucial dans le domaine des STI qui est celui de l’estimation. En effet, la mise en oeuvre delois de commande pour réguler le trafic impose de disposer de l’ensemble des informations concernantl’évolution de l’état du trafic. Dans ce contexte, deux algorithmes d’estimation sont proposés. Le premierrepose sur l’emploi du modèle METANET et les techniques de modes de glissement d’ordre supérieur. Lesecond est basé sur les CTM (Cell Transmission Models). Plusieurs études comparatives avec les filtresde Kalman sont proposées. La seconde contribution concerne la régulation du trafic. L’accent est mis surle contrôle d’accès isolé en utilisant les algorithmes issus du mode de glissement d’ordre supérieur. Cettecommande est enrichie en introduisant une commande intégrée combinant le contrôle d’accès et le routagedynamique. L’ensemble des résultats, validé par simulation, est ensuite comparé aux stratégies classiquesnotamment le contrôle d’accès avec l’algorithme ALINEA. La troisième contribution traite des problèmesde coordination. En effet, l’objectif est d’appliquer le principe de la commande prédictive pour contrôlerplusieurs rampes d’accès simultanément. L’ensemble des contributions ont été validées en utilisant desdonnées réelles issues en grande partie de mesures effectuées sur des autoroutes françaises. Les résultatsobtenus ont montré un gain substantiel en termes de performances tels que la diminution du trajet, dutemps d’attente, de la consommation énergétique, ainsi que l’augmentation de la vitesse moyenne. Cesrésultats permettent d’envisager plusieurs perspectives nouvelles de développement des recherches dansce domaine susceptibles d’apporter des solutions intéressantes. / The works presented in this PhD dissertation fit into the framework of Intelligent TransportationSystems. Although the beginnings of these systems have started since the 60s, their development, basedon information and communication technologies, has reached maturity during the early 80s. The ITS usesthe intelligence of different systems (embedded systems, intelligents sensors, intelligents highways, etc.)in order to optimize road infrastructures performances and respond to the daily problems of congestions.The dissertation presents four contributions into the framework of road traffic flow and tackles theestimation and control problems in order to eliminate or at least reduce the “recurrent" congestionsphenomena. The first point treats the problem of traffic state estimation which is of most importance inthe field of ITS. Indeed, the implementation and performance of any control strategy is closely relatedto the ability to have all needed information about the traffic state describing the dynamic behavior ofthe studied system. Two estimation algorithms are then proposed. The first one uses the “metanet"model and high order sliding mode techniques. The second is based on the so-called Cell TransmissionModels. Several comparative studies with the Kalman filters, which are the most used in road traffic flowengineering, are established in order to demonstrate the effectiveness of the proposed approaches. Thethree other contributions concern the problem of traffic flow control. At first, the focus is on the isolatedramp metering using an algorithm based on the high order sliding mode control. The second contributiondeals with the dynamic traffic routing problem based on the high order sliding mode control. Such controlstrategy is enriched by introducing the concept of integration, in the third contribution. Indeed, integratedcontrol consists of a combination of several traffic control algorithms. In this thesis the proposed approachcombines an algorithm of on-ramp control with a dynamic traffic routing control. The obtained results arevalidated via numerical simulations. The validated results of the proposed isolated ramp metering controlare compared with the most used ramp metering strategy : ALINEA. Finally, the last contributiontreats the coordination problems. The objective is to coordinate several ramps which cooperate andchange information in order to optimize the highway traffic flow and reduce the total travel time in theapplied area. All these contributions were validated using real data mostly from French freeways. Theobtained results show substantial gains in term of performances such as travel time, energetic consumptiondecreasing, as well as the increasing in the mean speed. These results allow to consider several furtherworks in order to provide more interesting and efficient solutions in the ITS field.

Les systèmes de transport intelligents ont acquis un grand intérêt pour la recherche ces dernières années. Alors que la simulation réaliste du trafic joue un rôle important, elle n'a pas reçu suffisamment d'attention. Cette thèse est consacrée à l'étude de la simulation du trafic au niveau microscopique et propose des modèles de mobilité des véhicules correspondants. À l'aide de méthodes d'apprentissage profond, ces modèles de mobilité ont fait leurs preuves avec une crédibilité prometteuse pour représenter les véhicules dans le monde réel. D'abord, un modèle de mobilité basé sur un réseau de neurones piloté par les données est proposé. Ce modèle provient de données de trajectoires du monde réel et permet de mimer des comportements de véhicules locaux. En analysant les performances de ce modèle de mobilité basé sur un apprentissage de base, nous indiquons qu’une amélioration est possible et proposons ses spécifications. Un MMC est alors introduit. La préparation de cette intégration est nécessaire, ce qui comprend un examen des modèles de mobilité traditionnels basés sur la dynamique et l’adaptation des modèles « classiques » à notre situation. Enfin, le modèle amélioré est présenté et une simulation de scénarios sophistiqués est construite pour valider les résultats théoriques. La performance de notre modèle de mobilité est prometteuse et des problèmes de mise en œuvre sont également discutés / The intelligent transportation systems gain great research interests in recent years. Although the realistic traffic simulation plays an important role, it has not received enough attention. This thesis is devoted to studying the traffic simulation in microscopic level, and proposes corresponding vehicular mobility models. Using deep learning methods, these mobility models have been proven with a promising credibility to represent the vehicles in real-world. Firstly, a data-driven neural network based mobility model is proposed. This model comes from real-world trajectory data and allows mimicking local vehicle behaviors. By analyzing the performance of this basic learning based mobility model, we indicate that an improvement is possible and we propose its specification. An HMM is then introduced. The preparation of this integration is necessary, which includes an examination of traditional dynamics based mobility models and the adaptation method of “classical” models to our situation. At last, the enhanced model is presented, and a sophisticated scenario simulation is built with it to validate the theoretical results. The performance of our mobility model is promising and implementation issues have also been discussed

La maintenance de flottes de systèmes mobiles dans le monde du transport et de la logistique revêt de nos jours une importance croissante de par l’augmentation des attentes des exploitants et opérateurs en termes de sécurité, de fiabilité, de suivi, de diagnostic et de maintenance de ces systèmes. Dans ce contexte, Les mainteneurs des flottes doivent souvent faire face à d'énormes quantités de données brutes, informations et événements de surveillance liés aux contexte de leurs systèmes. De plus, ces événements, données et informations manquent souvent de précision et sont souvent contradictoires ou obsolètes. Enfin, le degré d'urgence des décisions de maintenance est rarement pris en compte. Ce travail est consacré à la proposition et à la mise au point d’une architecture de surveillance pour l’aide à la maintenance d’une flotte de systèmes mobiles. Cette architecture, appelée EMH², est destinée à faciliter le diagnostic et le suivi de ce type de flotte. Elle est construite sur les principes holoniques, des plus bas (capteurs) aux plus hauts niveaux (ensemble d’une flotte de systèmes mobiles). Elle se base également sur une standardisation des événements traités afin de traiter les données de manière générique. Cette architecture, indépendante des types de systèmes surveillés et de leur niveau hiérarchique, peut devenir l'épine dorsale d’une stratégie efficace de maintenance proactive d’une flotte. Une méthodologie de déploiement est ainsi proposée. Une étude en simulation et une application sur une flotte de 10 trains actuellement en service est présentée. / The maintenance of mobile systems fleets in the world of transport and logistics is of increasing importance today due to the increasing expectations of operators in terms of safety, reliability, monitoring, diagnosis and maintenance of these systems. In this context, fleet maintainers often have to deal with huge amounts of raw data, information and monitoring events related to the context of their systems. Moreover, these events, data and information are often lack precision and often contradictory or obsolete. Finally, the urgency of maintenance decisions is rarely taken into account. This work is devoted to the proposal and the development of a monitoring architecture to help maintain a fleet of mobile systems. This architecture, called EMH², is intended to facilitate the diagnosis and monitoring of this type of fleet. It is built on holonic principles, from the lowest (sensors) to the highest levels (a whole fleet of mobile systems). It is also based on a standardization of processed events in order to process the data generically. This architecture, independent of the types of systems monitored and their hierarchical level, can become the backbone of an effective strategy for proactive fleet maintenance. A deployment methodology is thus proposed. A simulation study and an application on a fleet of 10 trains currently in service is presented.

In recent years, Intelligent Transportation Systems (ITS) have emerged as an efficient way of enhancing traffic flow, safety and management. These goals are realized by combining various technologies and analyzing the acquired data from vehicles and roadways. Among all ITS technologies, computer vision solutions have the advantages of high flexibility, easy maintenance and high price-performance ratio that make them very popular for transportation surveillance systems. However, computer vision solutions are demanding and challenging due to computational complexity, reliability, efficiency and accuracy among other aspects.   In this thesis, three transportation surveillance systems based on computer vision are presented. These systems are able to interpret the image data and extract the information about the presence, speed and class of vehicles, respectively. The image data in these proposed systems are acquired using Unmanned Aerial Vehicle (UAV) as a non-stationary source and roadside camera as a stationary source. The goal of these works is to enhance the general performance of accuracy and robustness of the systems with variant illumination and traffic conditions.   This is a compilation thesis in systems engineering consisting of three parts. The red thread through each part is a transportation surveillance system. The first part presents a change detection system using aerial images of a cargo port. The extracted information shows how the space is utilized at various times aiming for further management and development of the port. The proposed solution can be used at different viewpoints and illumination levels e.g. at sunset. The method is able to transform the images taken from different viewpoints and match them together. Thereafter, it detects discrepancies between the images using a proposed adaptive local threshold. In the second part, a video-based vehicle's speed estimation system is presented. The measured speeds are essential information for law enforcement and they also provide an estimation of traffic flow at certain points on the road. The system employs several intrusion lines to extract the movement pattern of each vehicle (non-equidistant sampling) as an input feature to the proposed analytical model. In addition, other parameters such as camera sampling rate and distances between intrusion lines are also taken into account to address the uncertainty in the measurements and to obtain the probability density function of the vehicle's speed. In the third part, a vehicle classification system is provided to categorize vehicles into \private car", \light trailer", \lorry or bus" and \heavy trailer". This information can be used by authorities for surveillance and development of the roads. The proposed system consists of multiple fuzzy c-means clusterings using input features of length, width and speed of each vehicle. The system has been constructed by using prior knowledge of traffic regulations regarding each class of vehicle in order to enhance the classification performance.

Indiana University-Purdue University Indianapolis (IUPUI) / With the rapid growth in wireless mobile communication technology, Vehicular Ad-hoc Network (VANET) has emerged as a promising method to effectively solve transportation-related issues. So far, most of researches on VANETs have been conducted with simulations as the real-world experiment is expensive. A core problem affecting the fidelity of simulation is the mobility model employed. In this thesis, a sophisticated traffic simulator capable of generating realistic vehicle traces is introduced. Combined with network simulator NS-2, we used this tool to evaluate the general performance of several routing protocols and studied the impact of intersections on simulation results. We show that static nodes near the intersection tend to become more active in packet delivery with higher transferred throughput.

Indiana University-Purdue University Indianapolis (IUPUI) / Pedestrian Automatic Emergency Braking (PAEB) system for avoiding/mitigating pedestrian crashes have been equipped on some passenger vehicles. At present, there are many e orts for the development of common standard for the performance evaluation of PAEB. The Transportation Active Safety Institute (TASI) at Indiana University-Purdue University-Indianapolis has been studying the problems and ad- dressing the concerns related to the establishment of such a standard with support from Toyota Collaborative Safety Research Center (CSRC). One of the important components in the PAEB evaluation is the development of standard testing facili- ties at night, in which 70% pedestrian crash social costs occurs [1]. The test facility should include representative low-illuminance environment to enable the examination of sensing and control functions of di erent PAEB systems. This thesis work focuses on modeling low-illuminance driving environment and describes an approach to recon- struct the lighting conditions. The goal of this research is to characterize and model light sources at a potential collision case at low-illuminance environment and deter- mine possible recreation of such environment for PAEB evaluation. This research is conducted in ve steps. The rst step is to identify lighting components that ap- pear frequently on a low-illuminance environment that a ect the performance of the PAEB. The identi ed lighting components include ambient light, same side/opposite side light poles, opposite side car headlight. Next step is to collect all potential pedes- trian collision cases at night with GPS coordinate information from TASI 110 CAR naturalistic driving study video database. Thirdly, since ambient lighting is relatively random and lack of a certain pattern, ambient light intensity for each potential col- lision case is de ned and processed as the average value of a region of interest on all video frames in this case. Fourth step is to classify interested light sources from the selected videos. The temporal pro le method, which compressing region of interest in video data (x,y,t) to image data (x,y), is introduced to scan certain prede ned region on the video. Due to the fact that light sources (except ambient light) impose distinct light patterns on the road, image patterns corresponding to speci c light sources can be recognized and classi ed. All light sources obtained are stamped with GPS coordinates and time information which are provided in corresponding data les along with the video. Lastly, by grouping all light source information of each repre- sentative street category, representative light description of each street category can be generated. Such light description can be used for lighting construction of PAEB test facility.

No description available.