Methoden zur Bewertung der Verbindungsqualität in Straßennetzen

The time required for a journey is the core indicator for describing the quality of services in a transport network. To use travel time data for planning and evaluation, travel time data from different sources must be comparable. If this is not guaranteed, identical or comparable traffic states might in reality be evaluated differently in planning depending on the data source. The objectives and the approach of the research project result from this requirement:
(1) Origin-destination travel times in road traffic were determined and compared using different methods and data sources (Google, HERE, INRIX, TomTom, travel demand model PTV-Validate). For this purpose, travel times of 16 selected connections were investigated;
(2) A generally accessible database for origin-destination travel times between central locations (metropolitan regions, regional centres, medium-sized centres) was created to assess time data from other sources. The database comprises 21,500 selected connections. For each connection, travel time data from three data sources (travel demand model, Google, TomTom) are provided for several times of day and percentiles;
(3) Based on the comparison of the methods and data sources, a concept was developed with which the time data of a particular data source selected by a user can be assessed using the developed database and if necessary corrected. In order to check the comparability of any analysis data source and a reference data source, a two-stage procedure is proposed consisting of a validation stage and an adaptation stage;
(4) Reference curves for assessing the quality of supply for the RIN (Guidelines for Integrated Network Design, RIN 2008) are suggested.
Three evaluation curves were estimated:
• Direct speed in the off-peak period;
• Direct speed in the peak hour without random disturbances;
• Direct speed in the peak hour with random disturbances.
The parameters of the reference curves were determined with two approaches. In the first approach, the empirical travel times of the database were used to estimate the parameters. In the second approach, travel times were derived from typical characteristics of the road network for the parameter estimation. The results should serve as a basis for further discussions on the revision of the RIN 2008 in the responsible committees.