Abstract This paper presents a model of traffic event streams processing. Events are generated by the developed spatiotemporal traffic simulator for real highway networks. The simulator is designed according to a distributed architecture based on mobile agents. It generates a flow of vehicles, assigning them to trips according to a model using geographic data. The highway network is equipped with sensors that generate events when passing vehicles. The event stream is processed in real time by agents to estimate the current traffic state to inform users via traffic message-variable panels. The architecture of the real-time event processing system is based on Kafka Stream Processing. To evaluate the performance of our model, we carried out a simulation of the traffic of a year in 24 h with a constraint of 25 million of Vehicles Kilometer per Day, producing an events density of 9485.2 Events/s. The proposed real-time processing topology shows that the estimation error does not exceed 5% for segments length less than 12 km.
An Efficient Traffic Monitoring Model Using a Stream Processing Platform Based on Smart Highways Events Generator
2018-05-15
10 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
TIBKAT | 1995
|Engineering Index Backfile | 1924
|Traffic monitoring platform system based on stream computing technology
Europäisches Patentamt | 2023
|Smart Vehicles Need Smart Highways
British Library Conference Proceedings | 1995
|Traffic efficiency of highways
Engineering Index Backfile | 1928
|