It may be stated that main explanatory variables in trip generation models are: number of inhabitants, number
of employed persons (in total and in particular sectors of economy) and number of trucks garaging in particular
TAZ (Traffic Analysis Zone). Moreover, main calculation methods used are multiple regression and trip
generation rates. Artificial neural networks (ANN) were used only for analysis of single generators (Al-Deek, H.
M., et al. (2005)).
Observing different approaches to freight modelling (vehicle/commodity based, I-O models) it may be seen
that the choice of approach will depend on data availability and model application. Even the most sophisticated
model will not provide good results when input data is uncertain or missing. Considering available data from
Polish surveys, vehicle based model was chosen. Trucks were divided into two groups: light (GVW less than 3.5
t) and heavy (GVW more than 3.5 t). When it comes to spatial division it was assumed that the region is
equivalent to province (there are 16 provinces in Poland) and TAZ is equivalent to commune.