ALL libraries (COBIB.SI union bibliographic/catalogue database)
  • Statistical modelling of traffic flow rate [Elektronski vir]
    Grabec, Igor ; Kalcher, Kurt, 1955 ; Švegl, Franc
    The road traffic is considered as a non-autonomous dynamic phenomenon and modelled statistically by a non-parametric approach. Information for the modelling is extracted from recorded time series of ... traffic flow rate. The time is represented by the hour and a day-code variable specified by the calendar. An optimal predictor of the traffic flow generator is formulated in terms of conditional average estimator. The condition is comprised of hour, day-code and several data of past flow rate. As an example the traffic flow rate at a representative point on a Slovenian high-way is modelled. Seasonal variation is taken into account by using in the modelling just one month long interval of the past flow rate record. The model is utilized to forecast traffic flow rate. Applicability of the method is evaluated by the correlationcoefficient r of the forecasted and original data. The mean value <r>~0.95 indicates rather good modelling. The performance of forecasting depends on the combination of variables representing the condition. The graph of r versus the number of condition components exhibits a maximum that determines an optimal combination of components. By the hour and day-code variables the mean traffic flow rate record over a week is determined. Its structure is described by a superposition of normal distributions whose parameters represent new information about the traffic phenomenon and activityof population.
    Type of material - conference contribution
    Publish date - 2008
    Language - english
    COBISS.SI-ID - 1409383