AUTOMATED REAL-TIME TRAFFIC FORECASTING SYSTEM
M. Dahmani
DOI: https://doi.org/10.33042/2522-1809-2022-4-171-76-81
2022-10-17
Municipal economy of cities
Abstract:Based on the analysis of the vehicles total number growth rates, which exceed the rates of expansion and optimization of the transport infrastructure, the need for the introduction of real-time traffic forecasting and control systems is shown. The factors that make it possible to detect the probability of potentially dangerous situations on the road, such as traffic jams, accidents and lack of parking spaces, respectively, in certain urban areas, based on the data of sensor networks and surveillance cameras combined within the global system of the Internet of Things, have been determined. It is proposed to build a sensor network based on magnetic sensors, which allows for high-precision geolocation with refinement of the received data by using ultrasonic sensors and optical monitoring tools, while identification is carried out by reading RFID tags. It is shown that the task of optimal organization of the relay system includes the determination of the features of the city infrastructure and statistical indicators of the city's traffic flows, and for a multi-level communication system, protocols are determined depending on the distance between nodes, requirements for the level of data protection, data transmission speed, minimum radio signal amplitude, as well as restrictions on the power supply of a separate node. The presented topology of the relay network includes the organization of sensor nodes into clusters, transmission from the main node of the cluster to the gateway node, and from the gateway nodes to the base station. On the basis of the specified model, a scheme for building cluster self-organization algorithms can be presented by forming clusters in real time according to the topology of the cluster tree, which allows simplifying the data transfer subsystem and reducing the time of processing input data. The developed scheme for analyzing the traffic flow at the intersection and the availability of parking spaces can be used in the development of methodological recommendations for the implementation of the "Smart City" concept and the creation of software applications that provide drivers with information about the state of traffic and predicted changes within a certain time interval. Keywords: automaticregulation of auto-traffic, controlled intersection, sensor network, Internet of Things, geolocation, identification, self-organization of clusters.