Distributed Learning Based Control for Multi Agent Systems
Multi-agent systems offer a great potential to improve the quality of modern society life. In the near future fleets of autonomous cars will be able to reduce traffic congestion and fuel consumption while increasing road safety. W...
ver más
¿Tienes un proyecto y buscas un partner? Gracias a nuestro motor inteligente podemos recomendarte los mejores socios y ponerte en contacto con ellos. Te lo explicamos en este video
Fecha límite de participación
Sin fecha límite de participación.
Descripción del proyecto
Multi-agent systems offer a great potential to improve the quality of modern society life. In the near future fleets of autonomous cars will be able to reduce traffic congestion and fuel consumption while increasing road safety. With almost half of all freight being transported by road, it makes up approximately a quarter of the total EU energy consumption and accounts for 18% of the greenhouse emissions. Fuel reduction in this area will have a significant impact on the environment. One way to achieve such reductions is through platooning where heavy-duty vehicles drive close to each other to reduce their aerodynamic drag and thus increase their fuel efficiency. While autonomous driving and platooning are areas of active research, open challenges arise in complex traffic scenarios with human interactions. Another challenge is that hierarchical control design with several different layers is required. The specific goals of the project are to develop novel algorithms for the control of safety-critical multi-agent systems in real-world scenarios, to understand the role of local informational constraints on the performance and safety of such systems and to design incentives for the individual agents that lead to a desired coordination of a fleet. This way global objectives will be optimized while accounting for complex traffic situations. The scientific contribution lies in combining and extending recent results from distributed predictive control, statistical learning and game theory as well as understanding the role of informational constraints in distributed learning-based control of multi-agent systems. The developed methods will have a high impact on both industry and society. In particular, the project will enable platooning in more complex scenarios, which has the potential to reduce fuel consumption of the transportation sector by up to 10% and thus make a significant contribution to the overall energy consumption and greenhouse emissions of the EU.