Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Macroscopic traffic flow modelling in the presence of vehicle automation and communication systems

Porfyri Kallirroi

Full record

---

Summary

Over the past few years most of the countries are facing the consequences of the ever-increasing number of vehicles, which lead to a continuous increase of congestion phenomena, resulting in significant increases in travel times, fuel consumption, and emissions, as well as reduced traffic safety. Conventional approaches for solving the problem of traffic congestion by expanding the existing infrastructure and operational improvements - such as auxiliary lanes, additional alternate routes and interchange modifications - remain practically infeasible, mainly due to economic and environmental reasons. Instead, comprehensive traffic control strategies can be defined to mitigate the problem of persisting traffic jams. However, the employment of efficient real-time traffic control measures entails the availability of reliable and robust traffic flow models that may be used to develop and validate the proposed control strategies. In this context, an effective calibration and validation process appears to be mandatory to ensure the credibility of traffic flow models in performing real-world simulations and optimization scenarios.Concurrently, engineers are seeking for solutions to improve the road network efficiency and capacity by means of Intelligent Transportation Systems (ITS). Specifically, during the last decade, an enormous continuing interdisciplinary effort is performed by the automobile industry, as well as by various research institutions around the world, to plan, develop, test, and start deploying a variety of Vehicle Automation and Communication Systems (VACS) that are expected to bring radical changes in the way the traffic flow will be controlled and optimized within the next decades. VACS, such as Adaptive Cruise Control (ACC) and Cooperative Adaptive Cruise Control (CACC) systems, have been initially developed based on new commuting alternatives for drivers and passengers, with particular emphasis given on improving comfort, convenience, and safety, as well as reducing traffic congestion.This thesis, which is composed of two parts, is an early attempt towards this direction. More specifically, the first part deals with the advancement of traffic flow models, with emphasis on macroscopic ones. In particular, within this thesis the well-known continuous second-order macroscopic gas-kinetic-based traffic model (GKT model) was validated, regarding the representation of traffic conditions at congested freeway areas. The model was calibrated and validated by employing an optimization methodology based on a parallel, metamodel-assisted Differential Evolution (DE) algorithm (synchronous and asynchronous) and using real traffic data from two different motorway networks; a motorway stretch in the U.K., where severe traffic congestion is created due to high on-ramp flows during the morning peak periods, and a freeway stretch in Greece, where recurrent congestion is triggered by a saturated off-ramp during the morning peak hours. Moreover, a multi-lane approach of the GKT model is evaluated using real traffic data from the aforementioned network in the U.K. Subsequently, by implementing the same optimization scheme, the GKT model was compared with the most popular discrete time-space macroscopic traffic flow model, namely the METANET model, in terms of the representation of traffic flow conditions at the motorway stretch in the U.K.The GKT second-order traffic flow model, presented in the first part of this dissertation, provides the methodical prerequisites for the second part of the dissertation, where a novel concept of two alternative models for the macroscopic simulation of ACC and CACC traffic is discussed. This approach is based on the introduction of a relaxation term in the momentum equation of the GKT model that satisfies the time/space-gap principle of ACC and CACC systems. In this thesis both linear and nonlinear stability analyses are performed, to derive the stability threshold of the aforementioned models, and additionally study the influence of the equipped vehicles on the traffic flow stabilization, with respect to both small and large perturbations around the equilibrium state.

Available Files

Services