Periodic Reporting for period 1 - CEGUM (Coordination for Efficient and Green Urban Mobility)
Période du rapport: 2019-06-01 au 2021-05-31
Nowadays, urban traffic congestion has become a permanent part of life in big cities. In addition to the negative environmental effects of traffic congestion and immense financial costs due to delays and the increase in fuel consumption, the life quality of individuals is influenced due to e.g. stress, and idle time behind red traffic lights. Congestion may also effect the safety by increasing the risk of accidents. As a remedy, methodological improvements along with advanced technological developments in communication systems, driver assistance systems, and also automated vehicles have turned the spotlight on more efficient use of the existing infrastructure. In the literature, this issue has been tackled independently from two different points of view, corresponding to two main players with a key role for mobility in urban networks: traffic lights and vehicles. Traffic signals serve the overall traffic flow efficiency while driver-assistance systems in vehicles serve the individual drivers’ interests. The aim of this project was to make traffic lights and vehicles work together and Coordinate for Efficient and Green Urban Mobility (CEGUM).
By offering a balanced trade-off between system- and user-optimum, CEGUM targets safer and faster mobility in cities with lower fuel consumption, less air pollution and less frustration for the European citizens due to reducing the idle-time behind the red lights.
By offering a balanced trade-off between system- and user-optimum, CEGUM targets safer and faster mobility in cities with lower fuel consumption, less air pollution and less frustration for the European citizens due to reducing the idle-time behind the red lights.
The main scientific contribution of this project so far is developing an appropriate control framework for coordinating both the traffic lights and the speed of the vehicles approaching an intersection in several movements. Such control framework should be able to take into account the computational complexity of the problem. Motivated by this fact, we proposed a decomposition technique to formulate the optimization problem.
The main achievement of the current stage of the project is this two-level agent-based solution. In the lower level, the trajectories of vehicles parametrised by the infrastructures’ timing will be optimized in each movement traveling toward the intersection. These parametrized trajectories will be fed to the upper level, where along with the constraints on the infrastructures’ scheduling, will be integrated into an optimization problem and solved using tools form parametric optimization. The product of the upper level, which is the optimized timing of the infrastructure, will be fed back to the lower level to determine the optimal trajectory of vehicles. With such decomposition, the coordination between infrastructure and vehicles moving in various directions is achieved while the computational complexity of the problem is addressed properly.
Simulation results confirm that having this coordination will reduce travel time and obviate the unnecessary acceleration and deceleration of vehicles which itself can save fuel consumption.
The main achievement of the current stage of the project is this two-level agent-based solution. In the lower level, the trajectories of vehicles parametrised by the infrastructures’ timing will be optimized in each movement traveling toward the intersection. These parametrized trajectories will be fed to the upper level, where along with the constraints on the infrastructures’ scheduling, will be integrated into an optimization problem and solved using tools form parametric optimization. The product of the upper level, which is the optimized timing of the infrastructure, will be fed back to the lower level to determine the optimal trajectory of vehicles. With such decomposition, the coordination between infrastructure and vehicles moving in various directions is achieved while the computational complexity of the problem is addressed properly.
Simulation results confirm that having this coordination will reduce travel time and obviate the unnecessary acceleration and deceleration of vehicles which itself can save fuel consumption.
Although only the first part of the proposal has been investigated due to the truncated duration of the project, the results are still very encouraging. It is a testimony to the importance of integrating vehicles and infrastructures together for a more efficient and green transportation in cities. Moreover, due to the urge and also competition for use of more and more sophisticated driver-assistance systems by all car-manufacturers around the world, the idea and results of CEGUM should be of great interest for this sector to have a negotiation with municipalities (or other entities involved) for making vehicles-infrastructure coordinations feasible. This change in thinking could have significant implications for current transport policies in cities.