โดย ผศ.ร.อ.ดร.

สุทธิพงษ์ มีใย ผูเ้ ชี่ยวชาญด้านแบบจาลองการขนส่งและจราจร

Overview of Model Structure

M-MAP2 | Samina.02 2
Network Model Development

M-MAP2 | Samina.02 3
Network Model: Level of Service

Network Overview

M-MAP2 | Samina.02 4
 Network Model: Level of Service Survey

Surveying to collect network attribute especially waiting time, travel time and transfer time

M-MAP2 | Samina.02 5
 Network Model: Level of Service

Example
MRT

Interarrival time
in each stop

M-MAP2 | Samina.02 6
 Network Model: Level of Service

Example
MRT Fare table

M-MAP2 | Samina.02 7
Horizontal and vertical transfer coding

Bang Wa station (S12)

Green Line

Simplify to
Horizontal transfer time = 2.74 min

Simplify to
Vertical transfer time = 1.31 min

Blue Line

Bang Wa station (BL34)

M-MAP2 | Samina.02 8
 Network Model: Level of Service Survey Transfer routes between stations (additional survey)

xxx = Silom Line

xxx = Blue Line
xxx = Purple Line
Bang Wa (S12) Bang Wa (BL34) Bang Wa (S12)
xxx = Gold Line
xxx = Red N-S Line
xxx = Red E-W Line

Krung Thon Buri (S7) Krung Thon Buri (G1) Krung Thon Buri (S7)

Bang Sue (BL11) Bang Sue (RN01) Bang Sue (BL11) Legend
Up
Down
Path
Latch
Bang Son (PP15) Bang Son (RW02) Bang Son (PP15)
Ticket Machine
Escalator
Stairs
Tha Phra (BL01) Platf. 1-2 Tha Phra (BL01) Platf. 1-2 Start/
Finish Point
Tha Phra (BL01) Platf. 3-4
M-MAP2 | Samina.02 9
 Network Model: Level of Service Survey Transfer situation between stations in the peak hours

Bang Son Bang Sue

Purple Line → Red Line : 6 (7) min Blue Line → Red Line: 6 (7) min
Red Line → Purple Line : 7 (7) min Red Line → Blue Line: 6 (6) min

Tao Poon Mo Chit / Chatuchak

Blue Line → Purple Line: 3 min Sukhumvit Line → Blue Line: 6 (7) min
Purple Line → Blue Line: 4 min Blue Line → Sukhumvit Line :7 (7) min

Siam Phaya Thai

Sukhumvit Line → Silom Line: 2 min Sukhumvit Line → ARL: 4 (7) min
Silom Line → Sukhumvit Line: 2 min ARL → Sukhumvit Line: 6 (6) min

Tha Phra
Platf.1-2 Blue Line → Platf.3-4 Blue Line: 1 min Makkasan / Phetchaburi
Platf.3-4 Blue Line → Platf.1-2 Blue Line: 1 min ARL → Blue Line : 9 (10) min
Blue Line → ARL : 9 (10) min
Bang Wa
Asok / Sukhumvit
Silom Line → Blue Line : 4 (4) min
Sukhumvit Line → Blue Line: 6 (7) min
Blue Line → Silom Line : 4 (5) min
Blue Line → Sukhumvit Line: 5 (5) min
Consultant survey on February 17, 2022 Krung Thon Buri
Silom Line → Gold Line : 4 (5) min Saladaeng / Silom
JICA team survey during 28-30 Nov 2017 Gold Line → Silom Line : 4 (5) min Silom Line → Blue Line: 11 (13) min
Blue Line → Silom Line: 10 (11) min
*The value in brackets is the transfer time using the ticket machine, *Platf. = Platform
M-MAP2 | Samina.02 10
Network Model: Level of Service

Example
Transfer station
Link: Centroid Connector

Node

Link: Expressway Link: Horizontal transfer

Link: Vertical transfer
Link: Station Access
Stop: MRT station

Centroid: TAZ
Link: Road

M-MAP2 | Samina.02 11
Network Model: Level of Service

Congestion survey

Note: Numbers 1 is survey date and time, 2 is head between trains (in minutes), 3 is
number of trains the passengers miss due to congestion, and 4 is congestion ratio
Source: M-MAP2 blueprint, JICA 2018

M-MAP2 | Samina.02 12
Network Model: Level of Service Survey The result of the travel time base on observed data

M-MAP2 | Samina.02 13
 Network Model: Level of Service

Example
Bus

M-MAP2 | Samina.02 14
Network Model: Level of Service Link type
Free-Flow Speed Capacity Mode
Link Type Details
(kph) (Vehicle/direction) Car Truck Mc Walk PT
1 Centroid connector (general) 15 0 Y Y Y Y Y
2 Centroid (PT + link-based highway assignment) 15 0 Y Y Y Y Y
3 Centroid connector to MRT 15 0 Y Y
4 Walk link 4.4 0 Y
5 Ferry link 15 0 Y
6 SRT rail link 35 0 Y
7 MRT rail link 35 0 Y
8 One way road 1 lane/direction 40 1,250 Y Y Y Y Y
9 One way road 2 lane/direction 40 2,500 Y Y Y Y Y
10 One way road 3 lane/direction 40 3,750 Y Y Y Y Y
11 One way road 4 lane/direction 40 5,000 Y Y Y Y Y
12 One way road 6 lane/direction 40 7,500 Y Y Y Y Y
13 Dual carriageways with frontage road 1 lane/direction 95 1,500 Y Y Y Y Y
14 Dual carriageways with frontage road 2 lane/direction 95 3,000 Y Y Y Y Y
15 Dual carriageways with frontage road 3 lane/direction 95 4,500 Y Y Y Y Y
16 Dual carriageways with frontage road 4 lane/direction 95 6,000 Y Y Y Y Y
17 Dual carriageways with frontage road 5 lane/direction 95 7,500 Y Y Y Y Y
18 Flyover 1 lane/direction 30 1,250 Y Y Y Y
19 Flyover 1 lane/direction 50 1,250 Y Y Y Y
20 Flyover 2 lane/direction 50 2,500 Y Y Y Y
21 Flyover 2 lane/direction 70 2,500 Y Y Y Y
22 Flyover 3 lane/direction 50 3,750 Y Y Y Y
M-MAP2 | Samina.02 15
Network Model: Level of Service Link type
Free-Flow Speed Capacity Mode
Link Type Details
(kph) (Vehicle/direction) Car Truck Mc Walk PT
23 Flyover 4 lane/direction 50 5,000 Y Y Y Y
24 Bus only link 1 lane/direction 40 1,200 Y Y
25 Minor road 1 lane/direction 30 1,200 Y Y Y Y Y
26 Minor road 1 lane/direction 40 800 Y Y Y Y Y
27 Minor road 1 lane/direction 60 1,000 Y Y Y Y Y
28 Minor road 1 lane/direction 60 1,200 Y Y Y Y Y
29 Minor road 2 lane/direction 60 2,400 Y Y Y Y Y
30 Minor road 3 lane/direction 60 3,600 Y Y Y Y Y
31 Major road 1 lane/direction 50 1,250 Y Y Y Y Y
32 Major road 2 lane/direction 50 2,500 Y Y Y Y Y
33 Major road 3 lane/direction 50 3,750 Y Y Y Y Y
34 Major road 4 lane/direction 50 5,000 Y Y Y Y Y
35 Major road 5 lane/direction 50 6,250 Y Y Y Y Y
36 Major road 6 lane/direction 50 7,500 Y Y Y Y Y
37 Major road 2 lane/direction 70 2,500 Y Y Y Y Y
38 Major road 3 lane/direction 70 3,750 Y Y Y Y Y
39 Major road 4 lane/direction 70 5,000 Y Y Y Y Y
40 Major road 5 lane/direction 70 6,250 Y Y Y Y Y
41 Major road 6 lane/direction 70 7,500 Y Y Y Y Y
42 Major road 3 lane/direction 95 3,750 Y Y Y Y Y
43 Major road 4 lane/direction 95 5,000 Y Y Y Y Y
44 Major road 5 lane/direction 95 6,250 Y Y Y Y Y
M-MAP2 | Samina.02 16
Network Model: Level of Service Link type

Free-Flow Speed Capacity Mode

Link Type Details
(kph) (Vehicle/direction) Car Truck Mc Walk PT
45 Major road 6 lane/direction 95 7,500 Y Y Y Y Y
46 Ramp expressway 1 lane/direction 40 1,800 Y Y Y
47 Ramp expressway 2 lane/direction 40 3,600 Y Y Y
48 Ramp expressway 3 lane/direction 40 5,400 Y Y Y
49 Expressways and tollway 1 lane/direction 95 1,800 Y Y Y
50 Expressways and tollway 2 lane/direction 95 3,600 Y Y Y
51 Expressways and tollway 3 lane/direction 95 5,400 Y Y Y
52 Expressways and tollway 4 lane/direction 95 7,200 Y Y Y
53 Expressways and tollway 5 lane/direction 95 9,000 Y Y Y
54 U-turn 1 lane/direction 25 600 Y Y Y Y Y
55 Special link for junction modelling 3 lane/direction 40 3,000 Y Y Y Y Y
56 Vertical Transfer link 4.4 0 Y
57 Horizontal Transfer link 4.4 0 Y
58 Rail Station Access link 25 0 Y Y Y Y Y
59 Rail Station Entry Exit_link 25 0 Y Y Y Y Y
60 Dummy Controid Connector 15 0 Y Y Y Y Y

M-MAP2 | Samina.02 17
Network Model: Level of Service Link type

Disable mode

M-MAP2 | Samina.02 18
Network Model: Level of Service

Average speed in CBD 18.52 km/hr (both direction)

Average speed in Urban 26.82 km/hr (both direction)
Average speed in Sub-urban 49.69 km/hr (both direction)

Model Car speed

M-MAP2 | Samina.02 19
 Network Model: Level of Service

Validation travel time

M-MAP2 | Samina.02 20
 Network Model: Level of Service

Validation travel time

M-MAP2 | Samina.02 21
Network Model: Level of Service

Dimensions
Purpose, Mode, Time, User,
Result and Iteration

M-MAP2 | Samina.02 22
Network Model: Level of Service Survey Parking Fee and Park&Ride survey

Parking Fee
1 Paragon
2 The Mall Bang Kae
3 The Mall Bang Kapi
4 Emporium
5 Central World
6 Central ladprao
7 Central Rama 9
8 Terminal 21 Asok
9 MBK Center

Park&Ride Fee
1 Lad Prao Station
2 Bang Phi Station
3 Lak Sog Station
4 Yak Kor Por Aor Station
5 Khaeha Station

M-MAP2 | Samina.02 23
Trip Generation Model
Development

24
 Trip Generation Model Development

• Using trip rate by trip purposes (age groups are considered in trip purposes)
𝑇𝑜𝑡𝑎𝑙 𝑡𝑟𝑖𝑝 𝑓𝑜𝑟 𝑝𝑢𝑟𝑝𝑜𝑠𝑒𝑖
𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑜𝑟 𝐴𝑡𝑡𝑟𝑎𝑐𝑡𝑖𝑜𝑛 𝑟𝑎𝑡𝑒𝑖 =
𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑖𝑛𝑑𝑒𝑥𝑖
• Trip purpose 𝑖:
1. Home to Work
2. Home to School
3. Home to Private unemployed
4. Home to Private WFH
5. Home to Private elder
6. Non Home-based Private
7. Work-Business
8. Work to Home
9. School to Home
10. Private to Home unemployed
11. Private to Home WFH
12. Private to Home elder

M-MAP2 | Samina.02 25
 Summary Trip Rate

Production (Origin) Attraction (Destination)

Trip purpose
Population Index Trip Rate Population Index Trip Rate
Home to work จำนวนประชำกรที่มีงำนทำ ช่วงอำยุ 15 – 60 ปี 0.998 จำนวนกำรจ้ำงงำน 0.990
Home to school จำนวนประชำกรที่เรียนช่วง อำยุ 5 – 21 ปี 0.995 จำนวนที่นั่งเรียน 0.888
Home to private unemployed จำนวนประชำกรที่อำยุ 22 – 60 ปี ที่ไม่มีงำนทำ 0.374 จำนวนกำรจ้ำงงำนภำคบริกำร 0.021
Home to private WFH จำนวนประชำกรที่อำยุ 22 – 60 ปี ที่ทำงำนที่บ้ำน 0.671 จำนวนกำรจ้ำงงำนภำคบริกำร 0.152
Home to private elder จำนวนประชำกรที่อำยุมำกกว่ำ 60 ปี 0.274 จำนวนกำรจ้ำงงำนภำคบริกำร 0.041
Non home-based to private จำนวนกำรจ้ำงงำน 0.039 จำนวนกำรจ้ำงงำนภำคบริกำร 0.064
Work-business จำนวนกำรจ้ำงงำน 0.055 จำนวนกำรจ้ำงงำน 0.055
Work to home จำนวนกำรจ้ำงงำน 0.972 จำนวนประชำกรที่มีงำนทำ ช่วงอำยุ 15 – 60 ปี 0.979
School to home จำนวนที่นั่งเรียน 0.842 จำนวนประชำกรที่เรียนช่วง อำยุ 5 – 21 ปี 0.944
Private to home unemployed จำนวนกำรจ้ำงงำนภำคบริกำร 0.026 จำนวนประชำกรที่อำยุ 22 – 60 ปี ที่ไม่มีงำนทำ 0.459
Private to home WFH จำนวนกำรจ้ำงงำนภำคบริกำร 0.172 จำนวนประชำกรที่อำยุ 22 – 60 ปี ที่ทำงำนที่บ้ำน 0.761
Private to home elder จำนวนกำรจ้ำงงำนภำคบริกำร 0.041 จำนวนประชำกรที่อำยุมำกกว่ำ 60 ปี 0.266
M-MAP2 | Samina.02 26
Summary Trip Matrix

Trip Matrix

Zone

Trip production 48 combinations and trip attraction 48 combinations (12 trip purposes X 4 vehicle availability)
M-MAP2 | Samina.02 27
Result of Trip Generation Base Year 2019

Trip purpose Trip/day

Home to work 10,808,456
Home to School 2,194,965
Home to private unemployed 142,958
Home to private WFH 1,025,931
Home to private elder 282,367
Non home-based private 422,741
Work-Business 596,173
Work to Home 10,535,998
School to Home 2,020,476
Private to Home unemployed 172,303
Private to Home WFH 1,139,850
Private to Home elder 265,081
Total 29,607,298
M-MAP2 | Samina.02 28
Trip Distribution Model
Development

29
 Trip Generation Model Development

• Using Gravity model by 12 trip purposes and 4 vehicle availability

• Trip purposes consist of: Vehicle availability consist of:
1. Home to Work
2. Home to School 1. No vehicle
3. Home to Private unemployed 2. Motorcycle only
4. Home to Private WFH
5. Home to Private elder 3. Passenger car only
6. Non Home-based Private 4. Multi vehicles (both motorcycle
7. Work-Business and passenger car)
8. Work to Home
9. School to Home
10. Private to Home unemployed
11. Private to Home WFH
12. Private to Home elder

M-MAP2 | Samina.02 30
 Gravity Model

▪ Gravity models:

𝑇𝑖𝑗 = 𝑎𝑖 𝑏𝑗 𝑃𝑖 𝐴𝑗 𝑓(𝑐𝑖𝑗 )

𝑇𝑖𝑗 = Trips from traffic zone 𝑖 to traffic zone 𝑗

𝑃𝑖 = Trip production of traffic zone 𝑖
𝐴𝑗 = Trip attraction of traffic zone 𝑗
𝑓 𝑐𝑖𝑗 = Deterrence function of generalized cost from traffic
zone 𝑖 to traffic zone 𝑗
𝑎𝑖 , 𝑏𝑗 = Adjustment factor of traffic zone 𝑖 and traffic zone 𝑗

M-MAP2 | Samina.02 31
 Deterrence Function

▪ Gamma Distribution:

𝑥 𝛼−1 𝑒 −𝛽𝑥 𝛽𝛼
𝑓 𝑥 = 𝑓𝑜𝑟 𝑥 > 0 𝛼, 𝛽 > 0
Γ 𝛼
Probability density function

Γ(𝛼) = Gamma function

𝛼, 𝛽 = Parameters
With a shape parameter 𝑘 and a scale parameter 𝜃
With a shape parameter 𝛼 = 𝑘 and an inverse scale
parameter 𝛽 = 1/𝜃, called a rate parameter
Source: https://en.wikipedia.org/wiki/Gamma_distribution

M-MAP2 | Samina.02 32
Deterrence Function
Vehicle
Trip Purpose Shape Rate Loglikelihood AIC BIC
Ownership
School to Home Non vehicle 2.997 0.049 -2,258 4,519 4,527
Motorcycle only 2.998 0.118 -4,145 8,294 8,304
Car only 2.976 0.131 -4,127 8,257 8,267
Multi vehicle 3.001 0.145 -9,050 18,103 18,115
Private to Home unemployed Non vehicle 3.818 0.060 -975 1,955 1,961
Motorcycle only 2.922 0.136 -1,750 3,505 3,513
Car only 2.566 0.122 -4,832 9,668 9,678
Multi vehicle 2.815 0.138 -5,734 11,473 11,483
Private to Home WFH Non vehicle 3.221 0.054 -6,560 13,125 13,135
Motorcycle only 2.883 0.120 -20,433 40,871 40,884
Car only 2.883 0.120 -31,348 62,699 62,713
Multi vehicle 2.760 0.122 -47,995 95,994 96,009
Private to Home elder Non vehicle 2.896 0.046 -3,858 7,721 7,730
Motorcycle only 2.358 0.099 -3,641 7,285 7,295
Car only 2.725 0.118 -10,695 21,395 21,407
Multi vehicle 2.542 0.117 -12,686 25,376 25,388
M-MAP2 | Samina.02 33
Deterrence Function
Vehicle
Trip Purpose Shape Rate Loglikelihood AIC BIC
Ownership
Home to Work Non vehicle 2.666 0.041 -12,432 24,868 24,880
Motorcycle only 3.290 0.124 -46,650 93,305 93,319
Car only 3.904 0.147 -52,302 104,609 104,624
Multi vehicle 3.519 0.140 -90,746 181,496 181,513
Home to School Non vehicle 2.988 0.048 -2,213 4,429 4,437
Motorcycle only 2.916 0.114 -4,043 8,090 8,100
Car only 3.059 0.135 -3,980 7,964 7,974
Multi vehicle 2.992 0.145 -9,007 18,017 18,029
Home to Private unemployed Non vehicle 3.681 0.057 -956 1,915 1,922
Motorcycle only 2.839 0.134 -1,710 3,423 3,432
Car only 2.532 0.122 -4,773 9,549 9,560
Multi vehicle 2.843 0.140 -5,705 11,414 11,424
Home to Private WFH Non vehicle 3.026 0.052 -5,076 10,155 10,165
Motorcycle only 2.772 0.119 -16,333 32,670 32,682
Car only 2.783 0.117 -26,806 53,616 53,630
Multi vehicle 2.673 0.119 -41,394 82,792 82,807
M-MAP2 | Samina.02 34
Deterrence Function
Vehicle
Trip Purpose Shape Rate Loglikelihood AIC BIC
Ownership
Home to Private elder Non vehicle 2.767 0.043 -3,789 7,582 7,592
Motorcycle only 2.314 0.098 -3,568 7,141 7,150
Car only 2.726 0.118 -10,563 21,129 21,141
Multi vehicle 2.507 0.115 -12,613 25,229 25,241
Non Home-based Private Non vehicle 3.621 0.062 -1,575 3,153 3,161
Motorcycle only 2.968 0.118 -3,098 6,200 6,210
Car only 2.948 0.150 -4,791 9,585 9,595
Multi vehicle 3.497 0.181 -5,779 11,562 11,573
Work-Business Non vehicle 3.482 0.054 -2,019 4,043 4,051
Motorcycle only 2.661 0.093 -4,238 8,479 8,489
Car only 3.154 0.128 -4,906 9,815 9,826
Multi vehicle 2.860 0.120 -5,316 10,635 10,646
Work to Home Non vehicle 2.643 0.041 -11,348 22,701 22,712
Motorcycle only 3.250 0.122 -43,354 86,711 86,726
Car only 3.915 0.147 -48,578 97,160 97,174
Multi vehicle 3.501 0.139 -84,537 169,079 169,095
M-MAP2 | Samina.02 35
 Graph Deterrence Function

Home to Work Home to School Home to Private Home to Private Home to Private Non Home-
unemployed WFH elder Based Private

Work to Home School to Home Private to Home Private to Home Private to Home Work-Business
unemployed WFH elder

M-MAP2 | Samina.02 36
Summary Trip Matrix

Trip Generation Trip Distribution

12 Trip Purpose
48 x OD Matrix

4 vehicle availability
M-MAP2 | Samina.02 37
Trip Distribution

Group District Level

(Gravity)

M-MAP2 | Samina.02 38
Station Access Model
Development

M-MAP2 | Samina.02 39
 Station Access Model

Station Access Model

Calculate the share of mode used for railway station
access from walk, motorcycle taxi, feeder and car Model Variables
Variables Detail
𝑒𝑥𝑝 𝑉𝑖
𝑃𝑖 = ASC Alternative specific constant for each mode. Walk/bike as reference
σ𝑛𝑗=1 𝑒𝑥𝑝 𝑉𝑗 Non-Motorized Travel (Walk only)
Time (Minute) Walk: walking time between origin or destination to station
𝑉𝑖 = ෍ 𝜃𝑘 ∙ 𝑋𝑖𝑘
𝑘
(Motorcycle taxi, feeder, car)
Motorcycle taxi: travel time and waiting time between origin or
Where: Motorized Travel Time
destination to station
𝑃𝑖 = Probability of selecting mode 𝑖 Feeder: travel time and waiting time between origin or destination to
(Minute)
𝑉𝑖 = Utility of selecting mode 𝑖 station, access/egress time to station included
Car: travel time between origin or destination to station, parking time
𝑛= Number of transport access/egress modes:
included
4 (walk, motorcycle taxi, feeder and car)
𝜃𝑘 = 𝑘𝑡ℎ utility parameter Travel Cost (Baht) Total cost spent between origin or destination to station
𝑋𝑖𝑘 = 𝑘𝑡ℎ utility variable of mode 𝑖 Waiting Time (Minutes) (Feeder, Motorcycle taxi) Waiting time for feeder bus or motorcycle taxi

M-MAP2 | Samina.02 40
 Station Access Model

Mixed Logit model

All Purpose Note: Random coefficient of travel time
Variables
assume normal distribution
parameter t-stat SD t-stat
WALK/BIKE - -
MC -2.266 -34.879
ASC - -
FEEDER -4.101 -35.381
CAR -2.802 -25.420
WALK/BIKE -0.091 -29.547 -3.332e-07 -1.5042e-04
Travel time (min)
Motorized -0.032 -5.419 -2.3160e-04 0.02220
Travel Cost (baht) Motorized -0.033 -11.323 - -
Number of obs. 10,094
Rho2 0.4346
AIC 15,841
BIC 15,898
Final Log Likelihood -7,912
WALK 4.5
Travel time VOT (baht/min)
Motorized 1.6
M-MAP2 | Samina.02 41
 Station Access Model

Example Log Sum = -1.972

Log Sum = -2.049
Log Sum Station Access
from zone 139

Log Sum = -1.944

Log Sum = -1.735

M-MAP2 | Samina.02 42
Modal Split Model
Development

M-MAP2 | Samina.02 43
 Main Modal Split

Details and assumptions

𝑛 𝑒𝑥𝑝 𝑉𝑖 1. Modal split model were constructed using RP data.
𝑃𝑖 = 𝐶𝑎𝑝𝑡𝑖𝑣𝑒𝑖 + (1 − ෍ 𝐶𝑎𝑝𝑡𝑖𝑣𝑒𝑗 ) 2. The model using mixed logit model with error components.
𝑗=1 σ𝑛𝑗=1 𝑒𝑥𝑝 𝑉𝑗 3. Assume normal distribution of coefficients for travel time, and error
components.
4. The non-captive part of the model was estimated from the non-captive
𝑉𝑖 = ෍ 𝜃𝑘 ∙ 𝑋𝑖𝑘 respondents only.
𝑘
Variables Detail
Where:
ASC Alternative specific constant for each mode. Car as reference
𝐶𝑎𝑝𝑡𝑖𝑣𝑒𝑖 = Probability of captive mode 𝑖
𝑃𝑖 = Probability of selecting mode 𝑖 Rail: boarding + transfer + waiting time
Bus and other public transports: boarding + waiting + access/egress to bus/other
𝑉𝑖 = Utility of selecting mode 𝑖 Travel Time (Minute)
public transports stop time
𝑛= Number of transport main modes: Car and motorcycle: driving + parking time
4 (rail, bus and other public transport,
Rail: rail fare
motorcycle, and car)
Bus and other public transports: bus and other public transports fare
𝜃𝑘 = 𝑘𝑡ℎ utility parameter Travel Cost (Baht)
Car: VOC (fuel, maintenance) + toll fee + parking fee
𝑋𝑖𝑘 = 𝑘𝑡ℎ utility variable of mode 𝑖 Motorcycle: VOC (fuel, maintenance) + parking fee
Short Trip Dummy (Car and Motorcycle) 1 for OD with distance less than 5 km, 0 otherwise;
Station Access (Rail alternative specific only) The log-sum of utility of access/egress modes

44
 Main Modal Split

Modal split model (RP data), using mixed logit model with error components
BMR Model
H-W + W-H H-S + S-H H-P + P-H (unemp) H-P + P-H (wfh) H-P + P-H (elder) OH-P + W-B
Mode Estimate t-stat Estimate t-stat Estimate t-stat Estimate t-stat Estimate t-stat Estimate t-stat
CAR
MC - - -1.817 -3.102 - - - - - - - -
ASC
BUS 1.440 3.188 0.777 1.447 2.487 2.454 1.364 2.333 2.480 2.452 1.467 3.235
RAIL 4.219 11.721 2.908 5.837 3.176 6.557 3.215 7.054 3.169 6.769 4.244 11.721
Travel time ALL -0.063 -3.919 -0.024 -2.410 -0.026 -3.829 -0.041 -3.484 -0.037 -3.858 -0.060 -3.966
Sigma Travel time ALL 0.030 2.602 - - - - - - 0.017 3.795 0.031 2.781
Travel cost ALL -0.018 -4.071 -0.016 -3.060 -0.013 -2.633 -0.015 -2.606 -0.013 -1.635 -0.019 -4.072
CAR 3.825 13.305 2.000 5.248 3.094 4.075 2.861 8.414 3.091 4.077 3.835 13.305
Vehicle Ownership
MC 2.372 9.835 2.420 5.285 1.088 2.206 2.391 7.392 1.085 2.214 2.381 9.852
CAR 1.494 4.537 0.582 1.332 1.158 2.349 1.788 4.226 1.155 2.362 1.509 4.564
Short Trip Dummy
MC 3.597 10.599 2.827 6.265 2.668 3.779 2.054 4.650 2.665 3.784 3.610 10.625
Station Access RAIL 0.802 6.541 0.338 3.168 0.519 3.310 0.727 3.920 0.518 3.312 0.811 6.525
Error Component 12 CAR, MC 1.065 2.216 - - - - - - - - 0.965 1.986
Error Component 4 RAIL 3.115 6.240 1.701 2.122 0.182 1.332 3.055 4.001 0.198 1.575 2.924 5.182

Number of obs. 2,549 1,406 897 1,341 118 211

Rho2 0.669 0.741 0.814 0.605 0.791 0.681
Start Log Likelihood -3168.900 -1706.740 -1077.810 -1670.190 -162.430 -276.730
Final Log Likelihood -1137.18 -499.00 -229.87 -1816.73 -22.04 -92.54
AIC 2300.36 1022.00 485.73 1460.71 91.05 204.04
BIC 2376.32 1084.99 548.12 1528.32 126.01 239.20
Travel time VOT 3.39 1.48 2.03 2.80 2.92 3.25
(baht/min)

Note: (1) assume random parameters as normal distribution for travel time, and error components,
(2) the table shows only significant coefficients (the coefficients do not significant in the model were removed).
M-MAP2 | EC.01 : Inception Report 45
Main Mode Choice Model

Different ASC Rail in each sector

Example
CBD (sector1, sector2)
- ASC rail = 1.0
Urban (sector3, sector4, sector5, sector6,sector7
sector8, sector9)
- ASC rail = 0.85
Suburban (sector10, sector11, ..)
- ASC rail = 0.25
Suburban red line (sector12, sector20, sector22)
- ASC rail = 0.10

M-MAP2 | Samina.02 46
Main Mode Choice Model

Captive
HH Vehicle Car_captive MC_captive Bus_captive
Non vehicle - - 9.75%
Motorcycle only - 17.43% -
Car only 45.01% - -
Multi vehicle 23.59% 16.84% -
Source: Main Mode Choice Survey

M-MAP2 | Samina.02 47
 Hub Access Model
Development

M-MAP2 | WOR.06 : Draft Final Report 48

 Hub Access Model
Data Collection about Air passenger
Suvarnabhumi Airport

Donmueang Airport

M-MAP2 | Samina.02 Source: AOT 49

 Hub Access Model
Data Collection about Intercity passenger
Bangkok Bus Terminal

Source: DLT

M-MAP2 | Samina.02 50
 Hub Access Model OD Hub Survey
Survey duration: 23 Jan 2023 – 19 May 2023
Number of Sample = 47,759

M-MAP2 | Samina.02 51
Hub Access Model OD Hub Survey

Gender and Age

N = 47,759
M-MAP2 | Samina.02 52
 Hub Access Model OD Hub Survey
Education level and Occupation

N = 47,759
M-MAP2 | Samina.02 53
 Hub Access Model OD Hub Survey
Income level

Mode Share

N = 47,759
M-MAP2 | Samina.02 54
 Hub Access Model OD Hub Survey

Suvarnabhumi
Airport

Mode Share

Nation

N = 18,010
M-MAP2 | Samina.02 55
 Hub Access Model OD Hub Survey

Donmueang
Airport

Mode Share

Nation

N = 11,408
M-MAP2 | Samina.02 56
 Hub Access Model OD Hub Survey

Bang Sue
Grand Station

Mode Share

Nation

N = 915
M-MAP2 | Samina.02 57
 Hub Access Model OD Hub Survey

Chatuchak
Bus Terminal

Mode Share

Nation

N = 9,707
M-MAP2 | Samina.02 58
 Hub Access Model OD Hub Survey

Ekkamai
Bus Terminal

Mode Share

Nation

N = 3,608
M-MAP2 | Samina.02 59
 Hub Access Model OD Hub Survey

New Sai Tai

Bus Terminal

Mode Share

Nation

N = 4,111
M-MAP2 | Samina.02 60
 Hub Access Model OD Hub Survey

Total (Expand)

Pass/day
Transport Hub (Base Year
2019)
Suvarnabhumi Airport 177,909
Donmueang Airport 113,078
Chatuchak Bus Terminal 92,538
Ekkamai Bus Terminal 27,235
New Sai Tai Bus Terminal 41,954

M-MAP2 | Samina.02 61
 Hub Access Model OD Hub Survey
Suvarnabhumi Donmueang Chatuchak
Airport Airport Bus Terminal

Ekkamai New Sai Tai

Bus Terminal Bus Terminal

M-MAP2 | Samina.02 62
 Hub Access Model Hub Access Survey
Thai Foreigner
Mode Estimate t-stat Estimate t-stat
Modal Split CAR

Hub Access Model ASC

MC
BUS
-1.928
-0.115
-6.068
-1.485 -6.507 -2.610
RAIL 0.763 2.327 -8.598 -2.672
Travel time ALL -0.073 -3.728 -0.149 -1.921
Sigma Travel time ALL 0.038 2.876 0.061 1.583
Travel cost ALL -0.016 -3.384 -0.018 -1.550
CAR 2.097 8.381
Vehicle Ownership
MC 0.712 2.801
CAR
Short Trip Dummy
MC 1.816 6.778
Station Access RAIL 1.091 4.475
Error Component 1 CAR 4.110 2.473
Error Component 4 RAIL 1.634 2.022

Number of obs. 1,161 610

2
Rho 0.4375 0.635
Start Log Likelihood -1638.55 -796.47
Final Log Likelihood -859.32 -232.14
AIC 1740.64 476.29
BIC 1796.27 502.77
Travel time VOT 4.38 8.13
(baht/min)

M-MAP2 | Samina.02 63
 Rail Route Assignment Model
Development

M-MAP2 | WOR.06 : Draft Final Report 64

 Rail Route Assignment

Rail Route Assignment

Calculate the route choice based on the mixed Model Variables
logit model: Variables Detail
Fare (Baht) Total route fare from origin station to destination station
𝑒𝑥𝑝 𝑉𝑖 Boarding Time (Minute) Total boarding time from origin station to destination station
𝑃𝑖 =
σ𝑛𝑗=1 𝑒𝑥𝑝 𝑉𝑗 Transfer Time, Horizontally (Minute) Time spent on walking horizontally while transferring the train
Time spent on moving vertically (elevator, escalator, etc.) while
𝑉𝑖 = ෍ 𝜃𝑘 ∙ 𝑋𝑖𝑘 Transfer Time, Vertically (Minute)
transferring the train
𝑘 Total time waiting for train boarding, including the first ride. Assume as
Waiting Time (Minute)
a half of headway.
Where:
Log-sum of the utility of every mode calculated railway station access
𝑃𝑖 = Probability of selecting route 𝑖 Station Access
model
𝑉𝑖 = Utility of selecting route 𝑖
Congestion Index Calculated from travel time and congestion rate, see below
𝑛= Number of candidate routes
𝜃𝑘 = 𝑘𝑡ℎ utility parameter Congestion Index Where:
𝑋𝑖𝑘 = 𝑘𝑡ℎ utility variable of route 𝑖 𝐶𝑅𝑚𝑗
2 𝐶𝐼𝑚 = Congestion index of route m
𝐶𝐼𝑚 = ෍ 𝐵𝑚𝑗 ∙ 𝐵𝑚𝑗 = Boarding Time of link j, route m
100
𝑗 𝐶𝑅𝑚𝑗 = Congestion rate of link j, route m
M-MAP2 | Samina.02 65
Rail Route Assignment

Variables
All purpose Mixed Logit model
parameter t-stat SD t-stat *Note: Random coefficient of time
Station Access (log-sum) 0.710 4.497 - - assume Normal distribution
Waiting time (min) -0.894 -6.057 0.639 7.343
In-vehicle time (min) -0.621 -3.796 0.887 4.652
Horizontal transfer time (min) -0.587 -3.121 0.400 3.365
Vertical transfer time (min) -0.651 -6.213 0.439 7.048
Congestion -1.508 -5.250 1.583 6.259
Fare (baht) -0.191 -17.273 - -
Number of obs. 12,456
Rho2 0.3154
AIC 3,466.81
BIC 3,563.40
Final Log Likelihood -1,720.4
In-vehicle 3.3
Horizontal Transfer 3.1
VOT (baht/min)
Vertical Transfer 3.4
Waiting 4.7

M-MAP2 | Samina.02 66
Model Validation

M-MAP2 | Samina.02 67
Model Validation

Mode Share
Source
Car Motorcycle Bus Rail
TDS survey 43.56% 29.53% 25.74% 1.18%
BTDS survey 52.61% 35.16% 9.73% 2.50%
Model (Base year) 48.48% 28.33% 20.80% 2.39%

M-MAP2 | Samina.02 68
 Model Validation

Level 1 : < 1 km.

Level 2 : 1 - 3 km.
Level 3 : 3 – 7 km.
Level 4 : > 7 km.
M-MAP2 | Samina.02 69
 Model Validation

Level 1 : < 1 km.

Level 2 : 1 - 3 km.
Level 3 : 3 – 7 km.
Level 4 : > 7 km.
M-MAP2 | Samina.02 70
 Model Validation
Case : Base Year
Boarding Passenger
Base Year 2019
Line (million trip-person/day)
Model Observe %diff
ARL 0.077 0.079 -2.5%
Purple 0.059 0.060 -1.7%
Green Sukhumvit 0.623 0.664 -6.2%
Green Silom 0.318 0.306 3.9%
Blue 0.337 0.322 4.7%
MAPE 3.8%

M-MAP2 | Samina.02 71