Mode Choice modeling using
Congestion Charging
Introduction
Objectives
Evaluate the perceived cost and actual
cost in congested situation
Formulate the Congestion Charging in
order to get the people to shift from
private car to LRT
Mode Choice Modeling
Finish
Mode Choice Modeling
Formulation of (Car LRT) Utility
function
Congestion Charging
Data
Steps in this Research
Data
Congestion
Utility Responden
Charging
function t
Data
Stated
Perceived Cost Preference :
Random
: Stated
Sampling
Questionnaire Preference
Questionnaire
Actual Cost :
Moving Car
Observer (10
times)
Bac
k
CONGESTION CHARGING
CC Formula
CC C MSC C MPC
m
ij
m
ij
m
ij
CCmij=Congestion Charging from i to j
using mode m
Cmij MSC = Marginal Social Cost from i to j
using mode m
Cmij MPC = Marginal Private Cost from i to
j using mode m
Flow vs Cost
Congestion Charging
=F-G
F = equilibrium
between demand and
supply (BCR=1)
Stubbs,
1980
Congestion charging = Actual
Generalized Cost
Generalized Cost
Vehicle operating cost ($/mile)
Both perceived and actual will be calculated
based on previous research method
Pollution Cost (Veh-mile)
idem
Travel time cost ($/min)
idem
Bac
k
Factors for Utility
function
Mode Choice affected by :
Travel Cost
Congestion Charging
Travel Time
Headway
Walking Time
Attribute Combination Survey
Design
(based on Cochran and Cox 1/3
Option from 25=32)
Combined Treatment
Attribute Level Difference
Travel Cost Congestion Cost Travel Time Headway Walking Time
1 (-) - - - - -
2 ab + + - - -
3 cd - - + + -
4 ace + - + - +
5 bce - + + - +
6 ade + - - + +
7 bde - + - + +
8 abcd + + + + -
Time Walking time
Travel Cost Congestion Travel Time
Headway to LRT Stop
Option ($) Cost ($) (min)
(min) (min)
Car LRT Car LRT Car LRT Car LRT Car LRT
1 6.00 5.00 4.00 - 10 16 - 15 - 8
2 7.00 5.00 6.00 - 10 16 - 15 - 8
3 6.00 5.00 4.00 - 12 14 - 12 - 8
4 7.00 5.00 4.00 - 12 14 - 15 - 5
5 6.00 5.00 6.00 - 12 14 - 15 - 5
6 7.00 5.00 4.00 - 10 16 - 12 - 5
7 6.00 5.00 6.00 - 10 16 - 12 - 5
8 7.00 5.00 6.00 - 12 14 - 12 - 8
Respond to attribute combination
(Dependent Variable for utility)
Number Semantic Scale Point Rating
1 Definitely choose car 1
2 Maybe choose car 2
3 Neutral 3
4 Maybe choose LRT 4
5 Definitely choose LRT 5
Numeric Scale
Point Probability Pr .Car
R ln
Rating (car) 1 Pr .Car
1 0.9 2.197
2 0.7 0.847
3 0.5 0.000
4 0.3 -0.847
5 0.1 -2.197
Utility Function
Uc ULRT = a + b1x1 + b2x2 + b3x3 +
b4x4 + b5x5
Uc ULRT = Utility of car minus utility of
LRT
x1 = Travel cost difference between car
and LRT
x2 = Congestion cost difference
between car and LRT
x3 = Travel time difference between
car and LRT
x4 = Headway difference between car
and LRT
Bac
x = Walking time difference betweenk
Mode Choice Modeling
Logit Binomial :
expU c exp U c U LRT
Pc
exp exp
Uc U LRT
1 exp U c U LRT
The probability of choosing the LRT :
1
PLRT 1 Pc
1 exp U c U LRT
OTHER METHOD : GAME
THEORY Bac
k
THANK YOU