Evaluation of Deterministic and Stochastic Components of Traffic Counts

Traffic counts or statistical evidence of the traffic processare often a characteristic of time-series data. In this paper fundamentalproblem of estimating deterministic and stochasticcomponents of a traffic process are considered, in the context of'generalised traffic modelling'. Different methods for identificationand/or elimination of the trend and seasonal componentsare applied for concrete traffic counts. Further investigationsand applications of ARIMA models, Hilbert space formulationsand state-space representations are suggested.