Using a regional model and real data from Maryland, researchers showed destination choice model to be better at predicting trip length and OD flows than gravity model with socioeconomic factors integrated
Introduction
Using the Maryland Statewide Transportation Model to demonstrate that destination choice model is a better method for trip distribution
Destination choice model is based on utility maximization
better at replicating HBW trips length and OD patterns
Concept
Trip distribution is the second step in the traditional 4-step model
It attempts to produce the best possible predictions of traveller’s destination choices on the basis of generation and attraction abilities of each zone and level of impedance between each OD pair
==Major weakness of the 4-step model because the interaction between origin and destination is complex and hard to capture==
Gravity model is the most widely accepted and used trip distribution model
Best suited at the aggregate level
Assumes trips produced at an origin and attracted to a destination are directly proportional to the total trip productions at the origin and total attractions at the destination
Random utility theory based models are often better suited at the disaggregate level but requires large amounts of data and is not widely applied in regional sized models
Incorporates travel time, socioeconomic variables
Requires individuals' demographic and socioeconomic attributes at the TAZ level which are often unavailable
Some researchers have incorporated or stratified the gravity model by socioeconomic factors
Adjusted friction factor to simulate complex interactions (non-linear) between origin and destination
There is a general lack of fitness comparison between the two models with real data and even less effort in comparing the two models within the context of a regional model
Results
The average trip length showed both models captured the general pattern, but it is an aggregate measure to determine which model is better
==The OD trips showed a better match for the destination choice model==