This paper describes how the availability of a comprehensive set of origin/ destination survey information has been used to undertake a comparison of assigned and observed routings and at the same time acquire an understanding of routeing behaviour and analyze the implications for traffic assignment techniques. This analysis demonstrates the extent of multi-routeing on an essentially inter-urban network and examines this multi-routeing behaviour in relation to network times and distances. It is argued that more consideration should be given to the modelling of movements on the correct routes with a current over-reliance on validation of volumes of link and screen-line flows. Also the need for novel assignment procedures for replicating route choice is identified. (A) This paper is also published in Seminar E, (Volume P307) p13-26 (see IRRD 842405).
Samenvatting