A thorough statistical approach to household‐based trip generation models is suggested. It involves the application of the Generalised Linear Modelling (GLM) framework to the selection of categories for Category Analysis. The example of analysis of long distance rail trip rates in Britain is used to demonstrate that the method is sufficiently versatile to handle data which follow an extremely skewed distribution. The trip rates were found to follow a negative binomial distribution which in turn comprised of a series of different Poisson distributions. This reflects the observation that the population can be divided into subgroups each with its own propensity to make trips. This observation was applied in conjunction with the GLM framework to give an appropriate model form. The paper concludes that the framework could also be applied to urban trip generation models, thus improving the reliability of urban transportation models.
Application of a generalised linear modelling approach to category analysis
Transportation Planning and Technology ; 13 , 2 ; 121-129
1989-02-01
9 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Unbekannt
An application of generalised linear modelling to the analysis of traffic accidents
Online Contents | 1996
|The generalised Frost-Dugdale approach to modelling fatigue crack growth
Tema Archiv | 2008
|Three-Dimensional Category Analysis Using Probabilistic Approach
British Library Online Contents | 1995
|Three-Dimensional Category Analysis Using Probabilistic Approach
Online Contents | 1995
|