Tra?c and transportation statistics are mainly published as aggregated information, and are traditionally based on surveys or secondary data sources, like public registers and companies’ administrations. Nowadays, advanced monitoring systems are installed in the road network, o?ering more abundant and detailed transport information than surveys and secondary data sources. Usually, these rich data are applied in the ?eld of transportation planning research. But they also seem promising to national statistics o?ces to update their databases and apply new methods to generate statistics. Transportation demand estimation and forecasting traffic volumes are taken as examples. Quantitative information on transportation demand is important for national and regional policy makers who want to know the number of freight vehicles traveling from origins to destinations. Traditionally, they largely extract this information from the national statistics o?ces. Transportation research needs the demand data to understand transportation behaviour in the road network, such as congestion and pollution. In the thesis, information methods and hierarchal Bayesian networks are used to demonstrate the approaches to estimate transportation demand. To forecast transportation demand, the hierarchal Bayesian network associated with the multi-process model is applied.
|Award date||3 Jun 2016|
|Place of Publication||Rotterdam|
|Publication status||Published - 3 Jun 2016|