Mode Choice

1
Mode Choice

• CE 573 Transportation Planning
• Lecture 15

2
Objectives

• mode split (choice) concept
• place in four-step-travel-demand-model (FSTDM)
• cause and effect
• model mode choice

3
What is mode split (choice)?

• Given
• an O-D volume between two zones,
• the set of modes available,
• the cost of using each of these modes
• How many travelers choose each of these modes?

4
Place in FSTDM
5
Mode Choice Cause and Effect

• How do these factors or variables affect your
  mode choice?
• Travel time
• Fare
• Wait-time
• Convenience
• Auto ownership
• Access time
• Availability (is a given mode even an option)

6
Mode Utility for a Traveler

• Quantify mode utility through a function
• Vm utility for mode m
• ?m,j parameter for mode m variable j

7
Probability of Choosing a Mode

• m index for modes
• M set of modes available
• Vm utility of mode m
• Pi probability of choosing mode i given all of
  the other modes and their utilities
• i index of the mode for which a probability is
  being estimated

8
Mode Choice Example Given Information
O-D Trip Matrix

• Mode choice from B to C
• Modes Auto and Bus
• Utility functions
• Where

9
Mode Choice Example Auto and Bus
10
Mode Choice Example Auto, Bus, and LRT
11
Discussion of error term ej in the context of IID

• Covariance of error terms i.e., COV(ebus, eLRT)
  ltgt 0.
• For a practical model, need to assume COV(ej, ei)
  0 for all j where j ltgt i.
• This is the assumption of IID which is defined as
  follows in the text

Where any two alternatives have the a non-zero
probability of being chosen, the ratio of one
probability over the other is unaffected by the
presence or absence of any additional alternative
in the choice set (Luce and Suppes, 1965).
12
Illustration of IID from two examples given above

• Observe ratio of Pauto/Pbus and how it changes
  regardless of the presence or absence of another
  alternative mode.
• Example 1 (Auto, Bus)
• Pauto/Pbus 0.677/0.323
• Example 2 (Auto, Bus, LRT)
• Pauto/Pbus 0.464/0.222