Session 38 Anahid Nabavi Larijani

22
13 January 2011 Evaluation of Real-time Control Strategies for Bus Line Number 1 in Stockholm by Anahid Nabavi Larijani Trafiksimulerin g Session 38

Transcript of Session 38 Anahid Nabavi Larijani

Page 1: Session 38 Anahid Nabavi Larijani

13 January 2011

Evaluation of Real-time Control Strategies for Bus Line

Number 1 in Stockholm

byAnahid Nabavi

Larijani

TrafiksimuleringSession 38

Page 2: Session 38 Anahid Nabavi Larijani

Outlines

Objective and motivation Target: blue bus line number 1 Holding control strategies Methodology: mesoscopic simulation model Results and analysis Conclusion

Page 3: Session 38 Anahid Nabavi Larijani

Objective and Motivation

• Unexpected interruption and uncertainties

Bus operat

ion

• Time table• Vehicle

assignment• Operation

management

Internal

factors

• Traffic jam• Bunching• Passenger

crowding

External

factors

As a solution: Holding Control Strategy

Trace bus movement at some points along the route Check the schedule association Adjust disturbances

Page 4: Session 38 Anahid Nabavi Larijani

Target bus line

blue bus line 1 afternoon peak – 15:30 to 18:00 high frequency – 4-6 minutes headway random arrival of passengers

Metro station

Tram station

Time point

Stora Essingen

Frihamnen

Page 5: Session 38 Anahid Nabavi Larijani

Bus line 1- Load profile

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 330

5

10

15

20

25

30

35

0

10

20

30

40

50

60

70

Boarding Alighting LoadStop

Nu

mb

er o

f Pas

sen

gers

Nu

mb

er o

f Pas

sen

gers

(A)

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 310

5

10

15

20

25

30

35

0

10

20

30

40

50

60

70

Boarding Alighting LoadStop

Nu

mb

er o

f Pas

sen

gers

Nu

mb

er o

f Pas

sen

gers

(B)

Page 6: Session 38 Anahid Nabavi Larijani

Holding criteria

Schedule-based holding strategy Headway-based holding strategy

minimum headway even headway

Holding locations – a set of time points

Schedule Based Minimum Headway Based Even Headway Based

3 time points:

current positionS1 MH1 EH1

4 time points:

proposed position S2 MH2 EH2

All stops are time

points: continues

control

S3 MH3 EH3

Page 7: Session 38 Anahid Nabavi Larijani

Methodology

Simulation model: BusMezzo

Input based of SL empirical data: Travel time distribution Vehicle scheduling Time-dependent demand Dwell time coefficients

Page 8: Session 38 Anahid Nabavi Larijani

Analysis

at most 8% error over 10 replications

Headway regularity – Service reliability Passenger time-cost Operator fleet certainty Driver relief point Level of service measures

Page 9: Session 38 Anahid Nabavi Larijani

Results-Service Reliability

Various strategies effect

1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829303132330

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

S1 MH1 EH1Stop

Coeffi

cie

nt

of

vari

ati

on o

f headw

ay

(EF33 : A)

1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627282930310

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

S1 MH1 EH1Stop

Coeffi

cie

nt

of

vari

ati

on o

f headw

ay

(FE31 : B)

Page 10: Session 38 Anahid Nabavi Larijani

Results-Service Reliability

Various sets of time points effect

Page 11: Session 38 Anahid Nabavi Larijani

Results-Headway regularity

0 60 120 180 240 300 360 420 480 540 600 More-5%

0%

5%

10%

15%

20%

25%

30%

35%

S1 MH1 EH1

Headway [sec.]

Fre

quency

0 60 120 180 240 300 360 420 480 540 600 More0%

10%

20%

30%

40%

50%

60%

70%

S3 MH3 EH3

Headway [sec.]

Fre

quency

bus bunching and schedule adherence

Page 12: Session 38 Anahid Nabavi Larijani

Results - Passenger time-cost

0 5 10 15 20 25 30 35 400

5

10

15

20

25

30

S1 MH1 EH1 S2 MH2 EH2 S3 MH3 EH3

Average excessed waiting time per passenger [sec.]

Ave

rage

incr

ease

in p

asse

nge

r in

-ve

hic

le t

ime

cau

sed

by

ho

ldin

g [s

ec.]

Trade-off between increase in in-vehicle travel time and waiting time

Page 13: Session 38 Anahid Nabavi Larijani

Results - Operator fleet certainty

90th percentile for vehicle assignment

5440 5560 5680 5800 5920 6040 6160 6280 6400 6520 More0%

5%

10%

15%

20%

25%

30%

S1 MH1 EH1Cycle time (total travel time)[sec.]

Freq

uen

cy

90th for S1 90th for MH1 95th for EH1

5440 5560 5680 5800 5920 6040 6160 6280 6400 6520 More0%

5%

10%

15%

20%

25%

30%

35%

40%

S2 MH2 EH2

Cycle time (total travel time) [sec.]

Freq

uen

cy

90th for S2 83th for MH2 99th for EH2

5440 5560 5680 5800 5920 6040 6160 6280 6400 6520 More0%

10%

20%

30%

40%

50%

60%

70%

S3 MH3 EH3

Cycle time (total travel time) [sec.]

Freq

uen

cy90th for S3 87th for MH3 98th for EH3

Page 14: Session 38 Anahid Nabavi Larijani

Results - Driver relief point

delay distribution at Fridhemsplan

<(-60)

(-60)-0

0-60 60-120

120-180

180-240

240-300

300-360

360-420

420-480

480-540

540-600

600<0%

5%

10%

15%

20%

25%

30%

S1 MH1 EH1Delay [sec.]

Freq

uen

cy

<(-60)

(-60)-0

0-60 60-120

120-180

180-240

240-300

300-360

360-420

420-480

480-540

540-600

600<0%

5%

10%

15%

20%

25%

30%

35%

S2 MH2 EH2

Delay [sec.]

Freq

uen

cy

<(-60)

(-60)-0

0-60 60-120

120-180

180-240

240-300

300-360

360-420

420-480

480-540

540-600

600<0%5%

10%15%20%25%30%35%40%45%

S3 MH3 EH3

Delay [sec.]

Freq

uen

cy

Page 15: Session 38 Anahid Nabavi Larijani

Results - Level of service measures

Scenario CV(h) Average waiting

time per

passenger

(sec)

Bunching

(%)

Average standing

time per

passenger

(sec)

On-time

arrivals

(%)

EF33 FE31 EF33 FE31

S1 0.39 0.54 172.94 7.91 83.09 79.62 79.24

S2 0.40 0.54 173.61 4.46 87.43 80.67 76.85

S3 0.36 0.50 165.34 5.96 97.30 86.90 80.05

MH1 0.36 0.39 159.96 2.23 83.33 62.99 69.79

MH2 0.36 0.37 158.42 2.00 81.78 61.02 67.42

MH3 0.24 0.26 146.50 0.49 89.14 65.22 71.42

EH1 0.26 0.35 151.35 1.58 67.12 58.41 78.66

EH2 0.24 0.31 147.38 0.87 69.20 56.35 76.65

EH3 0.16 0.18 141.04 0.27 85.93 58.46 67.29

Page 16: Session 38 Anahid Nabavi Larijani

• *Positive value stands for improvement

Results - Evaluation summary

Measures Strategy effect Time point layout effect

Headway reliability (CV) 40% 43%

Bunching 85% 62%

Relief point delay 11% in average with 41% lower variability

- 9% in average with24% lower variability

On-time performance - 0.5% - 3.7%

Total travel time (cycle) 5.3% 8.7%

Passenger waiting time 14% 6.5%

On-board standing time 24% -11%

Difference between S’s and EH’s scenarios under the same time point layout strategy effect Difference between 1’s and 3’s scenarios under the same holing strategy time point layout effect

Page 17: Session 38 Anahid Nabavi Larijani

Conclusion

The even headway-based holding strategy is a very promising bus control strategy since it yields to:

increase service reliability improve schedule adherence at relief point increase passenger comfortability decrease bus bunching and increase regularity fleet-cost saving for operators time-cost saving for passengers

Improved reliability

Higher efficiency

Higher LOS

Higher satisfaction

Page 18: Session 38 Anahid Nabavi Larijani

Thank You All

Page 19: Session 38 Anahid Nabavi Larijani
Page 20: Session 38 Anahid Nabavi Larijani

Target bus line-Operation background

Page 21: Session 38 Anahid Nabavi Larijani

Ontime52%

Delayed25%

Ahead

23%

Ontime Delayed Ahead

-10

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10111213141516171819200%

2%

4%

6%

8%

10%

12%

14%

16%

18%

20%

Delay [min.]

Frequency

Target bus line-Operation background

Page 22: Session 38 Anahid Nabavi Larijani

15:00-16:00 16:00-17:00 17:00-18:000.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

AverageTime Interval

Head

way

[min

.]

Sched-ule Head-way

(A)

Target bus line-Operation background