Abstract

 The RESILTRACK OBJECTIVE is to design a system which allows the knowledge of the infrastructure state and its affectation before climatic adverse real time phenomena, as well as prevent its behavior to act; where necessary, obtaining therefore railway infrastructures resilient before the effects of the climate change.

Document: Kick off CIMNE presentation

Abstract

In this paper, a comparative study between a hybrid technique that combines a Genetic Algorithm with a Cross Entropy method to optimize Fuzzy Rule-Based Systems, and literature techniques is presented. These techniques are applied to traffic congestion datasets in order to determine their performance in this area. Different types of datasets have been chosen. The used time horizons are 5, 15 and 30 min. Results show that the hybrid technique improves those results obtained by the techniques of the state of the art. In this way, the performed experimentation shows the competitiveness of the proposal in this area of application. Document type: Part of book or chapter of book

Abstract

O ônibus ainda é o modo de transporte público mais usado nas cidades brasileiras, mesmo aquelas que contam com sistemas sobre trilhos extensos. Já centenários, os sistemas de ônibus urbanos ainda se debatem na peleja diária da competição com os demais veículos e na busca de uma imagem mais favorável perante a seus usuários e à população. O advento do ferramental de Sistemas Inteligentes de Transporte (ITS) ofereceu a oportunidade de que esses sistemas pudessem contar com instrumentos de Planejamento, Programação, Monitoração e Controle Operacional próximos àqueles que os sistemas metroferroviários já dispõem há muitas décadas. Entre os vários avanços proporcionados pelo emprego de ITS no campo operacional, destacam-se as intervenções operacionais em tempo real, isto é, a aplicação de medidas que buscam oferecer ou restabelecer a regularidade e a eficiência da operação dos ônibus urbanos frente a contingências que eles enfrentam diariamente. Este trabalho objetiva fazer uma resenha das diversas opções de intervenções estudadas no campo acadêmico e executar simulações de três modalidades dessas estratégias. Foi construída uma rede reproduzindo um trecho de Corredor de ônibus na cidade de São Paulo e simulada a operação de uma linha de alta demanda, com base nos dados reais da operação, do trânsito (semáforos) e velocidades de percurso dos ônibus. A simulação, por sua vez, permitiu apurar os benefícios propiciados pelas intervenções testadas, chegando aos ganhos de tempos nas viagens e o possível aumento de oferta de veículos sem aumento da frota. Além de comparar os resultados entre os tipos de intervenção, foi verificada a sensibilidade dos resultados de cada estratégia a mudanças na intensidade de sua aplicação.Os resultados enquadram-se na faixa dos resultados obtidos nos diversos experimentos acadêmicos analisados. Dentro das expectativas, a intervenção das Meias Viagens (\\"deadheading\\") apresentou melhores resultados nos ganhos de tempos de viagem variando entre 8,5% e 12,9%, secundados pelas Paradas Limitadas (\\"skip-stop\\") variando entre 2,9% e 4,7%, enquanto a Linha Expressa apresentou resultados menores, motivados, principalmente, pelas limitações da extensão do trecho analisado. Buses are still the most common transit mode in Brazilian cities even in those that count with extensive rail systems. Century old city bus operations still struggle daily to compete with other road vehicles while trying to gain a better image before its users and the general public. The arrival of Intelligent Transportation System (ITS) solutions brought to these systems the opportunity of counting on Operational Planning, Programing, Monitoring and Control like those already used by rail transit for decades. Among several improvements possible using ITS real time intervention during operation stands out. This enables the application of measures to offer or restore regularity and efficiency of bus services in face of contingencies faced daily. This paper aims to review different options of intervention reported in academic publications and to perform simulations of three of these strategies. A network was modeled duplicating a real Sao Paulo bus corridor section on which a high demand bus line operation, based on actual operation data, traffic, traffic lights and bus speeds, was studied. The simulation, in turn, allowed us to assess the benefits of each alternative intervention tested by determining travel time gains and the increase in bus service supply without increasing the operating fleet. Besides comparing results between intervention types the sensitivity of each one\\'s results to strategy application intensity was also reached. The final experiments results fall within the range of those obtained in several other academic studies analyzed. Within the expectations, the deadheading intervention presented better results in travel time gains ranging from 8.5% to 12.9%, followed by the skip-stop, ranging from 2, 9% and 4.7%, while the Express Line presented lower results motivated, mainly, by the limitations of the extension of the analyzed section

Abstract

lotação do veículo é um dos fatores determinantes para a qualidade do transporte público (TP), pois, além do conforto do passageiro, afeta também a confiabilidade da linha, alterando a velocidade média do veículo. Entretanto, na cidade de São Paulo, Brasil, uma das maiores operações de ônibus do mundo, é difícil se encontrar implantado algum equipamento de contagem automática de passageiros (APC), provavelmente devido aos custos das tecnologias atuais. Esta pesquisa apresenta os resultados de um método de APC utilizando o sensoriamento por Wi-Fi de telefones móveis dos passageiros embarcados num veículo do TP por ônibus. Para este fim, um protótipo de baixo custo foi desenvolvido e instalado na linha 6500-10 entre o Terminal Santo Amaro e o Terminal Bandeira, durante 7 dias da semana em horários de pico e entre pico. Os dados coletados foram analisados através de ferramentas estatísticas e métodos de análise geográficas. Durante o experimento observou-se que o APC foi capaz de detectar a mudança da ocupação do veículo em 85% dos casos, com uma confiança de até 20% e um erro de até 22 passageiros por medida. Através de dados de posição geográfica foi possível gerar matrizes de carregamento, que permitem inferir a origem e o destino dos passageiros em cada viagem, assim como - para todo o período do experimento - indicar os trechos com maior lotação. Por outro lado, o método não apresentou bons resultados saindo dos terminais e quando a ocupação mudou repentinamente, requerendo estudos posteriores para refinar o algoritmo apresentado. Da mesma maneira, ainda não há como generalizar o método para toda operação, pois, o espaço amostral estudado foi pequeno, deixando como sugestão para pesquisas futuras o aprimoramento deste sistema. The crowdedness of a vehicle is one of the main quality factor of the public transportation (TP) quality, since, besides the comfort, also affect the reliability of the line, changing its average speed. However, the city of São Paulo, Brazil, one of the biggest bus operations of the world, it is difficult to find any automatic passenger counter (APC) equipment deployed, probably due the costs of current technologies. This research shows the results of an APC technology by the sensing of the Wi-Fi signal of smartphones that are boarded in a vehicle on TP by bus. Therefore, a prototype was developed and deployed on the line 6500-10 between the Terminal Santo Amaro and the Terminal Bandeira, during 7 working days of the week on peak and off-peak hours. was analyzed with statistical and geographical tools. During the experiment, it was observed that the Wi-Fi APC was capable of estimate the occupancy of the vehicle 85% of the cases, with a significance level up to 20% and an error up to 22 passengers per estimation. Through the geographical analysis it was generated load matrices, that allow to infer the origin and destination of the passenger in each trip, indicating the sections along the route where were more boarding and alighting. On the other hand, the method did not perform well when coming out of the final stops and when the crowdedness changed suddenly, requiring further analysis to understand how to deal better under these circumstances. Likewise, due to the limited sample utilized in this research, we cannot generalize the results presented, pressing the need for new researches on different operation scenarios.

Abstract

Orientador: Leandro Aparecido Villas Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação Made available in DSpace on 2018-08-30T12:25:57Z (GMT). No. of bitstreams: 1 Souza_AllanMarianode_M.pdf: 3090134 bytes, checksum: c9f97502249f7e0f959e5619ca372dcc (MD5) Previous issue date: 2016 Resumo: O congestionamento é um problema grave nas grandes cidades, o qual acaba gerando perdas econômicas, aumento do tempo de viagem e maior emissão de gases poluentes. Umas das principais causas do congestionamento é o aumento súbito do tráfego de veículos durante os horários de pico, acidentes de trânsito e gargalos nas infraestruturas de transporte. Nos últimos anos, com os avanços nas tecnologias de comunicação sem fio e o desenvolvimento de padrões para rede veiculares, tornou-se possível a implementação de Sistemas de Transporte Inteligentes. Os sistemas de transporte inteligentes são soluções abrangentes para gerenciamento de tráfego em tempo real que se baseia, sobretudo, em dados coletados de veículos, infraestruturas de rede disponíveis às margens de ruas, rodovias e de outros sensores que monitoram o trânsito. Uma área de pesquisa promissora para sistemas de transporte inteligentes são as redes veiculares, as quais são um tipo específico de redes móveis, nas quais os nós são formados por veículos com capacidade de processamento, armazenamento e comunicação sem fio. Com esta questão em mente, nessa dissertação são propostos dois sistemas de transporte inteligentes baseados em redes veiculares para detecção e controle de congestionamentos. Em particular, para controlar congestionamentos causados por acidentes de trânsito é proposto o SPARTAN, um sistema capaz de evitar a iminência de congestionamentos em locais de acidente de trânsito, o qual detecta o acidente e alerta os veículos próximos, para que eles alterem suas rotas e evitem passar pelo local do acidente, assim, evitando o surgimento de congestionamentos no local do acidente. Além disso, para controlar congestionamentos causados pelo grande número de veículos nas vias é proposto o COLOSSUS, um sistema capaz de classificar a condição de tráfego das vias, afim de detectar congestionamentos e, por fim, gerenciar o tráfego dos veículos, roteando-os para reduzir e controlar os congestionamentos detectados. Sendo assim, os resultados mostram a eficiência do SPARTAN para controlar congestionamentos causados por acidentes, uma vez que o mesmo reduz o tempo médio de viagem, tempo médio de congestionamento, consumo médio de combustível e emissões de CO em até 62%, 88%, 36% e 35% respectivamente. Por outro lado, a eficiência do COLOSSUS pode ser observada em seus resultados, onde o mesmo reduz o tempo médio de viagem, tempo médio de congestionamento, consumo médio de combustível e emissões de CO em até 34%, 70%, 12% e 11% respectivamente para congestionamentos causados pelo grande número de veículos Abstract: Congestion is a major problem in large cities, which ends up causing economic losses, increases the travel time and increases greenhouse gas emissions. One of the main causes of congestion is sudden increase in vehicle traffic during peak hours, traffic accidents and bottlenecks in transport infrastructure. In recent years, with advances in wireless communication technologies and the development of standards of vehicular ad hoc networks, has became possible the implementation of Intelligent Transportation Systems. Intelligent Transportation Systems are comprehensive solutions for real-time traffic management that is based primarily on data collected from vehicles, network infrastructure and other sensors that monitor the traffic. A promising research area in intelligent transportation systems are vehicular ad hoc networks, which are a specific type of mobile networks in which the nodes are formed by vehicles with processing power, storage, and wireless communication. With this question in mind, this work proposes two intelligent transport systems based on vehicle ad hoc networks for sensing and control congestions. In particular, to control congestion caused by traffic accidents is proposed the SPARTAN, which can avoid the rise of a congestion in accident areas, SPARTAN detects the accident and warns oncoming vehicles, so they change their routes and avoid to pass in the congestion area, thus avoiding the rise of a congestion in an accident area. Furthermore, to control congestion caused by the large number of vehicles on the roads is proposed COLOSSUS, a system that are able to classify the traffic condition in all roads in order to detect congestion and, finally, manage traffic of vehicles, routing them to reduce and control the detected congestion. Thus, the results of SPARTAN showed their effectiveness for controlling congestions caused by accident, since it reduces the average travel time, average congestion time, average fuel consumption and CO emissions by 62% 88% 36% and 35% respectively. On the other hand, the COLOSSUS efficiency can be seen in its results, where it reduces the average travel time, average time of congestion, average fuel consumption and CO emissions by up to 34% 70% 12% and 11% respectively for congestion caused by large number of vehicles on the roads Mestrado Ciência da Computação Mestre em Ciência da Computação 132901/2015-3 CNPQ

Abstract

Τα αυτοκινητιστικά δυστυχήματα αποτελούν μια από τις κυριότερες αιτίες θανάτου παγκοσμίως. Ο αυξανόμενος αριθμός τους οδήγησε στην συνειδητοποίηση ότι η χρήση προηγμένης τεχνολογίας για την κατασκευή ασφαλέστερων οχημάτων είναι απαραίτητη για την μείωση των ατυχημάτων και κατά συνέπεια των θανάτων που οφείλονται σε αυτά. Από τη στιγμή που οι τεχνολογικές εξελίξεις επέτρεψαν την ενσωμάτωση φθηνών, χαμηλής κατανάλωσης συστημάτων με μεγάλη επεξεργαστική ταχύτητα σε οχήματα, κατέστη προφανές ότι περίπλοκες τεχνικές υπολογιστικής όρασης μπορούσαν πλέον να χρησιμοποιηθούν για την υποβοήθηση της οδήγησης. Σε αυτή την κατεύθυνση, η παρούσα διατριβή εστιάζει στην ανάπτυξη καινοτόμων λύσεων για διαφορετικά κομμάτια που εμπλέκονται στα προηγμένα συστήματα υποβοήθησης του οδηγού. Πιο συγκεκριμένα, σε αυτή την διατριβή προτείνονται καινοτόμα υποσυστήματα για την αναγνώριση σημάτων οδικής κυκλοφορίας, την αναγνώριση φωτεινών σηματοδοτών, τον εντοπισμό προπορευόμενου οχήματος και τον εντοπισμό δρόμου. Οι τεχνικές που χρησιμοποιήθηκαν για την ανάπτυξη των προτεινόμενων λύσεων βασίζονται στην χρωματική επεξεργασία εικόνας με έμφαση στην ανεξαρτησία από την φωτεινότητα της σκηνής, στην χρήση πληροφορίας συμμετρίας για τον εντοπισμό χαρακτηριστικών αντικειμένων (όπως σήματα οδικής κυκλοφορίας, φωτεινοί σηματοδότες και οχήματα), στην χώρο-χρονική παρακολούθηση των εντοπισμένων αντικειμένων και στην αυτόματη κατάτμηση εικόνας για τον εντοπισμό δρόμου. Τα προτεινόμενα συστήματα αναπτύχθηκαν με στόχο την ανθεκτικότητα σε αλλαγές της φωτεινότητας ή τις καιρικές συνθήκες, καθώς και στην οδήγηση σε απαιτητικά περιβάλλοντα. Επίσης, έχει δοθεί ιδιαίτερη έμφαση στην προοπτική υλοποίησης συστημάτων πραγματικού χρόνου. Τα αποτελέσματα που παρουσιάζονται σε αυτή την διατριβή αποδεικνύουν την ανωτερότητα των προτεινόμενων μεθόδων έναντι αντίστοιχων της σχετικής βιβλιογραφίας, ειδικά στις περιπτώσεις του εντοπισμού προπορευόμενου οχήματος και του εντοπισμού δρόμου. Ελπίζουμε ότι μέρη της έρευνας αυτής θα εμπνεύσουν νέες προσεγγίσεις για τις μελλοντικές υλοποιήσεις αντίστοιχων συστημάτων. Traffic accidents are one of the main reasons for the loss of human lives worldwide. Their increasing number has led to the realization that the use of advanced technology for manufacturing safer vehicles is imperative for limiting casualties. Since technological breakthroughs allowed the incorporation of cheap, low consumption systems with high processing speeds in vehicles, it became apparent that complex computer vision techniques could be used to assist drivers in navigating their vehicles. In this direction, this thesis focuses on providing novel solutions for different tasks involved in advanced driver assistance systems. More specifically, this thesis proposes novel sub-systems for traffic sign recognition, traffic light recognition, preceding vehicle detection and road detection. The techniques used for developing the proposed solutions are based on color image processing with a focus on illumination invariance, using symmetry information for man-made objects (like traffic signs, traffic lights and vehicles) detection, spatiotemporal tracking of detected results and automated image segmentation for road detection. The proposed systems were implemented with a goal of robustness to changes of illumination and weather conditions, as well as to diverse driving environments. A special focus on the prospect for real-time implementation has also been given. The results presented in this thesis indicate the superiority of the proposed methods to their counterparts found in relevant literature in both normal and challenging conditions, especially in the cases of preceding vehicle detection and road detection. Hopefully, parts of this research will provide new insights for future developments in the field of intelligent transportation.

Abstract

Orientadores: Leandro Aparecido Villas, Edmundo Roberto Mauro Madeira Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Computação Made available in DSpace on 2020-02-07T18:33:19Z (GMT). No. of bitstreams: 1 Akabane_AdemarTakeo_D.pdf: 8120275 bytes, checksum: cd6482ee4f51bb13e31c71d5350d9fc7 (MD5) Previous issue date: 2019 Resumo: Devido à atual tendência mundial de urbanização, a sociedade moderna enfrenta, cada vez mais, sérios problemas de mobilidade urbana. Além disso, com o aumento constante do fluxo de tráfego veicular, as atuais soluções existentes para gerenciamento de tráfego se tornaram ineficientes. Com isso, para atender às crescentes necessidades dos sistemas de transporte, é necessário sistemas de transporte inteligentes (ITS). O desenvolvimento de ITS sustentável requer integração e interoperabilidade contínuas com tecnologias emergentes, tais como as redes veiculares (VANETs). As VANETs são consideradas uma tecnologia promissora que provê aplicações críticas de segurança e serviços de entretenimento, consequentemente melhorando a experiência de viagem do motorista e dos passageiros. Esta tese propõe um sistema de gerenciamento de tráfego de veículos sem a necessidade de uma infraestrutura de apoio. Para alcançar o sistema desejado foram necessários propor soluções intermediárias que contribuíram nesta tese. A primeira contribuição reside em uma solução que emprega conhecimento histórico dos padrões de mobilidade dos motoristas para obter uma visão global da situação da rede viária. Diferentemente de outras abordagens que precisam de troca constante de informações entre os veículos e o servidor central, nossa solução utiliza informações espaciais e temporais sobre padrões de mobilidade, além das informações específicas da infraestrutura viária, a fim de identificar congestionamentos no tráfego, permitindo, assim, o planejamento de roteamento de veículos. Como segunda contribuição, foi proposta uma solução distribuída para calcular a intermediação egocêntrica nas VANETs. Por meio da métrica egocêntrica foi proposto um mecanismo inovador de ranqueamento de veículos em redes altamente dinâmicas. As principais vantagens desse mecanismo para aplicações de VANETs são: (i) a redução do consumo de largura de banda e (ii) a superação do problema de topologias altamente dinâmicas. A terceira contribuição é uma solução de planejamento de rotas colaborativo com intuito de melhorar o gerenciamento do tráfego de veículos em cenários urbanos. Como última contribuição, esta tese integra as soluções descritas acima, propondo um sistema eficiente de gerenciamento de tráfego de veículos. As soluções propostas foram amplamente comparadas com outras soluções da literatura em diferentes métricas de avaliação de desempenho. Os resultados mostram que o sistema de gerenciamento de tráfego de veículos proposto é eficiente e escalável, qual pode ser uma boa alternativa para mitigar os problemas de mobilidade urbana Abstract: Due to the current global trend of urbanization, modern society is facing severe urban mobility problems. In addition, considering the constant increase in vehicular traffic on roads, existing traffic management solutions have become inefficient. In order to assist the increasing needs of transport systems today, there is a need for intelligent transportation systems (ITS). Developing a sustainable ITS requires seamless integration and interoperability with emerging technologies such as vehicular ad-hoc networks (VANETs). VANETs are considered to be a promising technology providing access to critical life-safety applications and infotainment services, consequently improving drivers¿ and passengers¿ on-road experiences. This thesis proposes an infrastructure-less vehicular traffic management system. To achieve such a system, intermediate solutions that contributed to this thesis were proposed. The first contribution lies in a solution that employs historical knowledge of driver mobility patterns to gain an overall view of the road network situation. Unlike other approaches that need constant information exchange between vehicles and the central server, our solution uses space and temporal information about mobility patterns, as well as road infrastructure information, in order to identify traffic congestion, thus allowing for vehicle routing planning. Secondly, a distributed solution to calculate egocentric betweenness in VANETs was proposed. Through the egocentric metric, an innovative vehicle ranking mechanism in highly dynamic networks was proposed. The main advantages of this mechanism for VANETs applications are (i) reduced bandwidth consumption and (ii) overcoming the problem of highly dynamic topologies. The third contribution is a collaborative route planning solution designed to improve vehicle traffic management in urban settings. As the last contribution, this thesis integrates the solutions described above, proposing an efficient vehicle traffic management system. The proposed solutions were widely compared with other literature solutions on different performance evaluation metrics. The evaluation results show that the proposed vehicle traffic management system is efficient, scalable, and cost-effective, which may be a good alternative to mitigate urban mobility problems Doutorado Ciência da Computação Doutor em Ciência da Computação 2015/25588-6 FAPESP

Abstract

This paper discusses the problem of modelling and simulation of traffic systems, and presents the traffic simulator SITS (Simulator of Intelligent Transportation Systems). The SITS is based on a microscopic simulation approach considering different types of vehicles, drivers and roads. A dynamical analysis of several traffic phenomena is then addressed. info:eu-repo/semantics/publishedVersion

Abstract

This paper presents the Simulator of Intelligent Transportation Systems (SITS). The SITS is based on a microscopic simulation approach to reproduce real traffic conditions in an urban or non-urban network and considers different type of vehicles, drivers and roads. A dynamical analysis of several traffic phenomena is then addressed. The results of using classical system theory tools point out that it is possible to study traffic systems, taking advantage on the knowledge gathered with automatic control algorithms. In this line of thought, it was also presented a new modelling formalism based on the embedding of statistics and Fourier transform. info:eu-repo/semantics/publishedVersion