In multi-agent systems, greedly agents can harm the performance of the overall system. This is the case of traffic commuting scenarios: drivers repete their actions trying to adapt to daily changes. In this domain, there are several proposals to achieve the traffic network equilibrium. Recently, the focus has shifted to information provision in several forms as a way to balance the load. Most of these works make strong assumptions such as the traffic authority and/or drivers having perfect information. In reality, the information the central control provides to drivers contains estimation errors. The goal of this paper is to propose a socially efficient load balance by internalizing social costs caused by agents' actions. Two issues are addressed: the model of information provision accounts for information imperfectness, and the equilibrium which emerges out of drivers route choices is close to the system optimum due to mechanisms of road pricing. The model can then be used for traffic authorities to simulate the effects of information provision and toll charging.

No abstract

No abstract

Developing a valid agent-based simulation model is not always straight forward, but involves a lot of prototyping, testing and analyzing until the right low-level behavior is fully specified and calibrated. Our aim is to replace the try and error search of a modeler by adaptive agents which learn a behavior that then can serve as a source of inspiration for the modeler. In this contribution, we suggest to use genetic programming as the learning mechanism. For this aim we developed a genetic programming framework integrated into the visual agent-based modeling and simulation tool SeSAm, providing similar easy-to-use functionality.

No abstract