The majority of freight transport models are either heuristic, transferring methodological basis from passenger transport models or macro-economic-like models. Rather, this project has developed a theoretical micro-economic methodological basis for freight transport based on stochastic utility theory. The basic methodological problem has not yet been solved satisfactorily in practice which is illustrated by a very limited number of international publications on freight models compared to passenger transport models.
Furthermore, the project has incorporated a detailed decision model and an optimisation model from the Centre's other projects into the freight transport model.
The purpose of the project is to develop the methodological basis for a new type of transport model to describe freight transport, and to test this in practice. The idea is to develop a freight transport model that describes the underlying decisions by producers and consumers and conveyors rather than adding on to heuristic methods e.g. the gravitation model or simpler adaptations of passenger transport models. The projects main aim is the development of the methodological basis.
Sub-activities of the project resemble the classical economic split between supply and demand which are moderated into goods and transport at a number of levels in the model system.
1. The demand model describes (1) the equilibrium between producer and consumer trade flows. This corresponds to passenger transport models' trip generation models, though with incorporation of more complex optimisations by use of utility theory. Based on stochastic utility theory a micro economic foundation is developed to facilitate modelling of each producer and consumers decision, including differences in their decision criteria, incomplete knowledge of the choice alternatives, and statistical uncertainties. (2) Choice of conveyor resembles the choice of mode in passenger transport modelling. For freight transport the combined choice of transport mode and conveyor by means of nested or cross-nested (logit) models, where the underlying utility functions are expected to be fundamentally different from those applied for passenger transport. (3)The overall demand is generated by the socio-economic model (project 8) on the contrary to passenger traffic models more simple trip rate models.
2. The supply model differs fundamentally from passenger transport models as the logistic optimisation models are incorporated directly into the supply estimation. Stochastic variation is added to preferences (Random Coefficients) and in general, within the optimisation routine. This corresponds to the development in mixed logit models for passenger transport, although the mathematical forms are more complex.
3. The network models reflect the result of the supply model and may therefore be decomposed into trips which can be modelled by a modified stochastic user equilibrium model, resembling previous work at CTT.
A methodological basis have been laid out for a model system, which can model effects of traffic political initiatives, such as investments in infrastructure, changes of tax and duty, subsiding and tax deductions and exemptions etc.
On the demand model side, the equilibrium between producer and consumer trade flows has been described in a SCGE (Spatial Computable Generalised Equilibrium) model. This model is then linked to models for decision concerning transport and logistics.
On the supply side, logistic optimisation models have been incorporated directly into the supply estimation which encompasses decision on a 3-level model layout (decisions made by transport buyers, transport organisers and by transport level, respectively). Stochastic variation is added to preferences and in general.
On the network side, the network models reflect the result of the supply model and may therefore be decomposed into trips, which can be modelled by a modified stochastic user equilibrium model.
The project has close ties to other modelling projects within CLG and other research programmes. During the project, a strong international research network has been established. Several models for freight transport have been investigated.