Seoul Development Institute Building a TDM Impact Analysis System for the Introduction of a Short-Term Congestion Management Program in Seoul Jin-Ki Eom,

Seoul Development Institute

Building a TDM Impact Analysis System for the Introduction of a Short-Term

Congestion Management Program in Seoul

Jin-Ki Eom, Kee-Yeon Hwang, Ikki KimResearcher of Seoul Development Institute (SDI)

San4-5, Yejang-dong, Jung-ku, Seoul 100-250, KoreaE-mail : [email protected]


SCMP(Sort-term Congestion Management Program)

SECOMM(Seoul Congestion Management Model)

SECOMM Case Study

Conclusion

Outline


Seoul has been known for the notoriety of its severe traffic congestion. In order to mitigate the congestion problems, the transportation policy of Seoul Metropolitan Government had been mainly focused on the supply of transportation systems until the early 1990’s.

The sharp decrease of investment on transportation infrastructures followed by recent economic recession.

The massive implementation of subway system does not reduce auto rider-ship as much as we expected.

Why Seoul needs SCMP ?

Short-term Congestion Management Program


1. Setting up Short-term Target of Traffic Management

2. Selecting TDM Programs to Reduce the Excessive Auto Demand

3. Building a Methodology for Forecasting the Expected Impacts of Programs(ex. SECOMM)

4. Monitoring Traffic Conditions Regularly

2. SCMP (Short-term Congestion Management Program)


Assumption

Structure of the SECOMM

Mode Split Model

Assignment Model

Link Travel Speed Adjustment Function

3. SECOMM(SEoul COngestion Management Model)


Assumption

The assumptions of SECOMM are as follows

Mode split and route choice are variable while trip generation and trip distribution are not in short-run

Investment is fixed in the short-run


Structure of the SECOMM

EMME/2

Macro

Short-term TrafficManagement Target

Household TravelSurvey

Data Clearance

END

If Satisfy the goalof Strategy

Yes

No

Modal Split Model(A-Logit)

Travel SpeedSurvey Data

in Seoul

Setting TDMAlternatives

ForecastingProspective

ImplementAlternatives

Monitoring/Analysis


Data Requirements

IndexData

FormatDate Source

’96 Seoul Metropolitan O/D by Modes- Peak Hour and Non-Peak Hour Day/hour

1997.10 –1998.3

SeoulDevelopment

InstituteTrip/TravelPattern ’96 Household Travel Diary Survey

- Travel Preference and Individual 1 weekTrip Data (Wednesday)

Day1997.10 –

1998.3

SeoulDevelopment

Institute

NetworkSeoul Metropolitan Road Network Data

Transit Route and Network DataHour

1996.11 –1998.4

SeoulDevelopment

Institute

A Day Traffic Volume at Namsan #1.3Tunnel

Hour1996.11 –

1998.4

SeoulMetropolitanGovernmentTraffic

VolumeBus & Subway Ridership Data &

Traffic Volume Survey Data in SeoulHour

1996,1997

SeoulMetropolitan

Police

Speed Traffic Speed Survey Data in Seoul Hour1996,1997

SeoulMetropolitanGovernment


Process of Building Mode Split Model

Variable Selection

Data Clearing

Mode Split of Cleared Data Set

Generating Unknown TravelTime & Cost

of Alternative Travel Modes

Building Data File(Convert Data into A-Logit Format)

Building Control File1: MNL Method / 2 : Nested Method

A-Logit Run

Satisfy Stopping Criteria

Yes

Utility Functions and Parameters are Set

VariablesAdjustment

No


U i : utility function of mode i Var i : constant(except bus) of mode i Ttime i : total travel time(min) of mode i TCost i : total travel cost(won) of mode i 21 : Coefficient of independent variable

Nested Tree for Each Alternative

Logit Model (1)

Car TaxiSubwayBus Car SubwayBusTaxi

Car TaxiSubwayBus Car TaxiSubwayBus

iii TCostTTimeVarUi 21


The Parameter Values and T-Values of Nested-Logit Models

VariableValue

Cost TimeAuto

DummySubwayDummy

TaxiDummy

1 2 2_2

ParameterValue

-0.000175 -0.03417 -0.6845 -0.8317 -2.211 0.9065 -

T-Value -5.6 -21.9 -5.9 -14.3 -8.7 21.6 -0.2434 0.1078

ParameterValue

-0.00023 -0.03988 -0.9568 -0.9200 -1.979 0.5084 0.3897

T-Value -5.0 -20.9 -4.1 -14.1 -6.4 7.9 5.10.2489 0.1143

ParameterValue

0.000083 -0.03753 0.0717 -0.8515 -1.7600 0.0445 11.61

T-Value 2.8 -21.8 0.3 -14.2 -8.4 2.8 9.90.2469 0.1119

ParameterValue

-0.000124 -0.03472 -0.6525 -0.8361 -2.173 0.8391 -

T-Value -3.3 -21.8 -5.5 -14.3 -9.1 16.8 -0.2440 0.1085

Logit Model (2)


Assignment Process

Trip IndexTrip Index Pre-Network

Build

Pre-NetworkBuild

HighwayNetwork

HighwayNetwork

Transit Network- Bus, Subway

Transit Network- Bus, Subway

Management Region- Link, Zone

Management Region- Link, Zone

Link Group- Speed, Volume

Link Group- Speed, Volume

TrafficManagement

Index Confirm

TrafficManagement

Index Confirm

`96 Traffic Census - Peak hour O/D - 1 Day O/D

`96 Traffic Census - Peak hour O/D - 1 Day O/D

Vehicle Occupancyfor Each Mode

(PCU )

Vehicle Occupancyfor Each Mode

(PCU )

Highway ODHighway OD

HighwayAssignment

HighwayAssignment

Assignment Result forEach Link

(Volume, Speed, Time)

Assignment Result forEach Link

(Volume, Speed, Time)

To Compare PresentTo Compare Present

NoNo

All Mode OD(Peak Hour / 1 Day)

All Mode OD(Peak Hour / 1 Day)

Auto ODAuto OD

Taxi ODTaxi OD

Bus ODBus OD

Subway ODSubway OD

HighwayNetwork

HighwayNetwork

LinkPerformance

Function

LinkPerformance

Function

Yes

S T O PS T O P

TransitNetwork

TransitNetwork

Transit ODTransit OD

TransitAssignment

TransitAssignment

To ApplyWeightFactor

(Wait timeAccess time,Boarding/Alighting

time)

To ApplyWeightFactor

(Wait timeAccess time,Boarding/Alighting

time)Assignment Result

for Each Route(Travel Time,Cost)

Assignment Result for Each Route

(Travel Time,Cost)

Traffic IndexAdjustment

Model Calibration


Using the adjustment factor, We can predict link travel speed

Process of Predicting Link Travel Speed

fS

SestL

obsL fSS est

lprel

Where, obsLS:Observed Average Speed of Link Group L estLS : Estimated Average Speed of Link Group L to Use

Estimated Link Volume from Assignment Result prelS :Predicted Average Speed of Link l f:Adjustment Factor


Study Process

Structure of Emme/2 Macro

Study Results

4. SECOMM Case Study

Study Title : Impact Analysis of Gasoline Tax Increase


Study Process

Building Mode SplitModel

Variable ParameterInput

1. Gasoline Tax Increase

2. New Mode Split Ratio Calculation UsingParameters of Mode Split Model

Network CalibrationBefore Tax Increse

Network CalibrationResults Saving

3. Rebuilding Trip O/D ofEach Mode

Auto-Assignment Transit-Assignment

Saving AutomobileTravel Time

Saving TransitTravel Time

5. New Mode Split Ratio Calculation UsingParameters of Mode Split Model

Adjust factor

6.Saving Difference of ModeSplit Ratio

7.If Satisfy StopingConditions

S T O P

No

Yes

4. Run Assignment

0. Initiation


Structure of Emme/2 Macro

Emme/2 Macro TREEEmme/2 Macro TREE

1) InitializedMacro

SubmacroSubmacro

SubmacroSub-submacro

1)2)

3)4)...

0) Main Macro

2) AssigmentMacro

4) Mode SplitCalibration

Macro

3) Time/CostCalclation

Macro

5) Rebuild TripMacro


The Speed Changes by Gasoline Tax Increase

Average Speed (km/h)

Mileage Tax Charging

$0.1 / l l Increase $0.2 / l l Increase $0.3 / l l IncreaseCategory Current

SpeedSpeed Change Rate Speed Change Rate Speed Change Rate

CBD 20.18 20.36 +0.2 0.9 20.39 +0.2 1.0 20.61 +0.4 2.1

Arterial 23.97 24.21 +0.2 1.0 24.34 +0.4 1.5 24.64 +0.7 2.8

UrbanHighway

42.54 43.08 +0.5 1.3 43.53 +1.0 2.3 44.18 +1.6 3.9

TotalAverageIn Seoul

25.19 25.33 +0.1 0.6 25.72 +0.5 2.1 26.10 +0.9 3.6

Case Study Results (1)


$0.2/L

18.5%

19.0%

19.5%

20.0%

20.5%

21.0%

0 1 2 3 4 5 6iteration

Before

After (the first stage)

After (the stable stage)

$0.1/L

$0.3/L

Auto-Mode Split Ratio Changes Resulting from Gasoline Tax Increase



<Response of SPEED to OIL_P>

Monitoring Data

- 300

- 200

- 100

0

100

200

0 5 10 15 20 25 30 35 40 45 50 55 60- 0.4

- 0.2

0

0.2

0.4

0.6

0 5 10 15 20 25 30 35 40 45 50 55 60

<Response of Car to OIL_P>

Response to Oil Price Increased


Peak Hour Auto Volume Changes Resulted by Gasoline Tax Increase

Before$0.1 / l lIncrease

$0.2 / l lIncrease

$0.3 / l lIncrease

Peak HourAutomobile

Volume(Vehicle/hour)

275,491 270,493 268,329 262,420

DecreasingVolume

(%)-

-4,998(-1.8)

-7,162(-2.6)

-13,071(-4.8)



SECOMM is a TDM impacts analysis system integrating mode choice model and trip assignment model in a module and iterating the interactions between them until the stop conditions are accomplished.

Using SECOMM, we can quickly forecast the impacts of TDM therefore, we can implement SCMP in Seoul.

To enhance the usefulness of SECOMM, there are several things to be done:

checking the estimated results of SECOMM through continuous monitoring on traffic situation in Seoul

updating the O-D data at least every 5 years

updating the network and travel behavior data

5. Conclusions

Seoul Development Institute Building a TDM Impact Analysis System for the Introduction of a Short-Term Congestion Management Program in Seoul Jin-Ki Eom,

Documents

secomm slide

seoul jinki eom

secomm case study study

oil price increased

mode choice model

trip assignment model

congestion problems

assumptions of secomm