Prototypical Sample Enumeration As a Basis for Forecasting with Disaggregate Models



Prototypical Sample Enumeration As a Basis for Forecasting with Disaggregate Models

Authors

DALY A J, Hague Consulting Group, The Netherlands

Description

The objective of disaggregate modelling as applied to travel demand forecasting is to explain the choices made by individual travellers. This approach has proved very successful as a basis for the development of models and through the technique of sample

Abstract

The objective of disaggregate modelling as applied to travel demand forecasting is to explain the choices made by individual travellers. This approach has proved very successful as a basis for the development of models and through the technique of sample enumeration disaggregate models have also been used successfully for short- term forecasting. However, longer-term forecasting is not possible with the straightforward applications of sample enumeration. To fill this gap, the technique of prototypical sample enumeration has been developed.

The function of prototypical sample enumeration is to provide a basis for the application of disaggregate travel demand models. Additionally, it allows the changing nature of the population to be taken into account in forecasting. Given the substantial changes that are currently taking place in the structure of Western European populations, such as reductions in household size, increasing employment of women and a general 'greying' of the population, the resulting changes in travel demand can also be quite large. Vrolijk et aI. (1987) predict a 19% growth in car kilometres over a 27-year period simply from changes in the population (including 6% overall growth), quite separately from increases in car ownership and licence holding. Thus predictions of changed population structure contribute to the forecasting of traffic changes in themselves, in addition to providing a basis for the application of other models.

The first section of this paper introduces the background of disaggregate modelling and sample enumeration in which prototypical sampling is applied. The second section states simply the background and main objectives of the method. Section 3 explains the details of the two principal methods that are used in practice to achieve those objectives: Iterative Proportional Fitting and Quadratic Optimisation. Some examples from practical studies are given in Section 4 and the paper concludes with a discussion of the main advantages and disadvantages of the method.

Publisher

Association for European Transport