AUTOMATED REAL-TIME TRAFFIC FORECASTING SYSTEM

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