WP3: Simulation of bicyclist behaviour

Coordinated by Hans Skov-Petersen, KU

As a synthesis of the knowledge gained during WP1 and WP2, an agent-based simulation model of individual bicyclists is constructed to assess the present and future flow of bicycles through the urban fabric. A main output of the model will be the spatial and temporal distribution of bicycles. In addition to the knowledge-base of WP1 and 2, information required for the WP will be obtained from a Stated Preference choice experiment coordinated with a GPS survey; the National Travel Survey; the 100×100m square net (population and workplaces); and municipality GIS-data. The technical development of the model will be the core of the Ph.D. study involved in the WP. The WP consists of three sub-modules:

WP3.1 Cycling trip generation: A probabilistic model of cycling trip generation based on population characteristics and bike-ability. The module will enable estimation of the number of bicycle trips (distributed over different mobility styles) that are generated from a location or zone during – for instance – a day. WP3.2 Configuration of cycling trips: A probabilistic model of trip types in terms of destinations (e.g. work or education) and possible intermediate stops (e.g. kinder gardens and/or shops) for different mobility styles. The module will enable estimation of trip configurations, given a location (of the trips origin) and mobility style. WP3.3 Cyclists’ route choices: A set of parameters enabling probabilistic choices between optional segments at network junctions, given mobility style, destination/stops and segment attributes e.g. safety, traffic load, pollution, route efficiency, aesthetics, and type (e.g. with/without bicycle lane). Choice parameters will be revealed from the itineraries registered by GPS (200 individuals will be registered). For individual itineraries, attributes of the network segments actually chosen at junctions are analysed, relative to the attributes of the segment that were deselected. The attributes include those applied in the Stated Preference choice experiment, enabling comparisons of the revealed and stated preferences. The cross-validation between the results of the two methods will strengthen conclusions and further qualify future application of Stated Preference choice experiments for ex-ante studies in relation to bicycle transport. WP3.4 Model programming and application: combining the former three sub-modules, a full model is complied and applied to cases in the municipality of Odense and at Vestvolden (Copenhagen municipality and adjacent municipalities) as part of the cycle-project ‘The fortification of Copenhagen’.