Abstract:
This paper presents a new multi agent based approach applying computational
intelligence concepts to implement a cooperative and well coordinated, Multiagent
system for real-time traffic control and management of a complex traffic network. The
large-scale traffic network is divided into various sub problems, and each sub problem
is handled by an intelligent agent with fuzzy neural and Case based reasoning based
decision-making module.Proposed solution help in quick incident diagnosis,
minimizing response delays by managing them and speed up the traffic incident
recovery to avoid the congestion. Key to effective response to any incident requires
extensive communication and institutional coordination .Ontology is used as the base
component for coordination and collaboration to address the research issues of efficient
coordination and consensus development among multiple autonomous agents acting as
robots or virtual cops at various intersections for decision making. The architecture
counts on several AI techniques germane to efficient decision making i.e. CBR & FNN.
The proposed architecture shows a high degree of adaptability leading to the least need
for human intervention. The proposed solution for traffic management is followed by a
case study that reveals the performance and effectiveness of the proposed architecture
over previous architectures.
