Research of Prognostic Abilities of Local Model of Controlled Process for Traffic Forecasting
暂无分享,去创建一个
Abstract The problem of traffic forecasting is analyzed. It is shown that the characteristic feature of transport flows is their non-stationarity. Prognostication of traffic by synthesis of local model of controlled process (LMCP) is proposed and the forecasting method is presented. The advantages of LMCP are the simplicity of mathematical apparatus and the possibility of organizing control in conditions of complexity, uncertainty and non-stationarity of the controlled process. The active accumulation of information about the controlled process in the process of LMCP synthesizing allows to reduce the problem of forecasting with known input and output and unknown external influences to the problem of forecasting with known input and output and absence of external influences, regardless of their number. Therefore, the offered method allows real time traffic forecasting, without the need for preliminary accumulation, processing and analysis of large amounts of data. The obtained results can be used to solution of practical problems arising in traffic control, in particular, to increase the throughput of highways, to reduce congestions and accelerate transport flows.
[1] Eleni I. Vlahogianni,et al. Short‐term traffic forecasting: Overview of objectives and methods , 2004 .
[2] Mohsen Guizani,et al. When Traffic Flow Prediction and Wireless Big Data Analytics Meet , 2019, IEEE Network.
[3] Eleni I. Vlahogianni,et al. Road Traffic Forecasting: Recent Advances and New Challenges , 2018, IEEE Intelligent Transportation Systems Magazine.