Gilberto Shigueo Nakamiti

Fuzzy distributed artificial intelligence systems

In this paper, we introduce a fuzzy distributed artificial intelligence (FDAI) approach. Through the use of fuzzy logic, we have been able to develop mechanisms that we feel may effectively improve current DAI systems, giving much more flexibility and providing the subsidies which a formal theory can bring. The appropriateness of the FDAI approach is explored in an important application, a fuzzy distributed traffic-light control system, where we have been able to aggregate and study several issues concerned with fuzzy and distributed artificial intelligence. We also present a number of current research directions necessary to develop the FDAI approach more fully.<<ETX>>

Fuzzy sets in distributed traffic control

The task of controlling urban traffic requires flexibility, adaptability and handling uncertain information spread through the intersection network. The use of fuzzy sets concepts convey these characteristics to improve system performance. This paper reviews a distributed traffic control system built upon a fuzzy distributed architecture previously developed by the authors. The emphasis of the paper is on the application of the system to control part of Campinas downtown area. Simulation experiments considering several traffic scenarios were performed to verify the capabilities of the system in controlling a set of coupled intersections. The performance of the proposed system is compared with conventional traffic control strategies under the same scenarios. The results obtained show that the distributed traffic control system outperforms conventional systems as far as average queues, average delay and maximum delay measures are concerned.

Intelligent Real-Time Traffic Control

The task of controlling urban traffic requires flexibility, adaptability, and handling of uncertain information spread through the urban network. The use of fuzzy sets, case-based mechanisms, and genetic algorithms in a distributed system framework can provide the necessary foundations for an efficient and adaptable on-line traffic control system. In this paper we present a distributed traffic control system built upon a fuzzy, adaptive framework. A traffic network consisting of 16 traffic lights in six intersections was implemented and compared with conventional traffic control approaches. Simulation results show that the distributed system outperforms current traffic engineering strategies.