Intelligent ontological agent for traffic light control of isolated intersections

As a result of increasing parameters and different conditions in traffic light control domain, modeling and updating this domain enables traffic agents to understand the environment better and make decisions effectively. By considering this viewpoint, firstly, we propose a fuzzy traffic light ontology which is made manually based on traffic standards and expert's knowledge. Next, an intelligent agent is designed to apply the ontology as a knowledge base to conclude observed traffic data and control isolated intersections. The ontology has been evaluated by experts; also we have simulated this system and compared it with other traffic light control systems. Results show the proposed system reduces average delayed time of each vehicle at each cycle more than fixed time and vehicle actuated approach, effectively.