Ali Gholami Center for Advanced Transportation Education and Research (CATER)

2014 ITE Intermountain Section Annual MeetingCenter for Advanced Transportation Education and ResearchUniversity of Nevada, Reno

Using Detector Information to Determine Turning Movement Proportions at Signalized

Intersections with Shared Lanes

Ali GholamiCenter for Advanced Transportation Education and Research (CATER)

Department of Civil & Environmental EngineeringUniversity of Nevada, Reno

Center for Advanced Transportation Education and ResearchUniversity of Nevada, Reno

Project Panel MeetingMarch 27, 10122014 ITE Intermountain Section Annual MeetingCenter for Advanced Transportation

Education and ResearchUniversity of Nevada, Reno

Outline

Volume application

Problem statement

Proposed methods

Case study intersections

Intersection operational analyses Traffic safety studies Travel demand modeling Identifying critical flow time periods …

Turning Movement Volume Applications

Loop Detector4

Loop Detector5

What is the current practice?6

Loop Detector Data , Aggregated7

9 10 11 12 5 6 7 8

29 33 45 41 0 0 0 0

26 36 39 37 0 0 0 0

33 32 57 53 0 0 0 0

46 67 89 87 0 0 0 0. . . . . . . . . .. . . . . . . . . .; ; ; ; ; ; ; ; ; ;

6/12/2013 6:15:00AM

6/12/2013 6:30:00AM

6/12/2013 6:45:00AM

Volume Report

Detectors

Date/ Time

KIETZKE & MOANA

6/12/2013 6:00:00AM

Loop Detector Data, High Resolution8

Sample data collected from the data communication adaptor Data Communication Adaptor

•Expensive•Time assuming•Not continuous

•Manual

•Not possible in shared lanes

•Automatic using loops

Collecting intersection Volume in Nevada

• Only high resolution data

• Both stop bar and departure detectors are necessary

• Departure detector

• Aggregated or high resolution data• Network

Equilibrium (NE)

• 1 min aggregated data or high resolution data

• Volume and Queue Length of Shared Lanes

• Only high resolution data•Flow Characteristics of Shared Lanes (FC)

• Aggregated or high resolution data •Observed proportions in the field (OP)

Determine Turning Movement Proportions at Signalized Intersections with Shared Lanes

Network Equilibrium (NE)11

𝑊𝑇 𝑖𝑡=−𝑁𝐿𝑖

𝑡−𝑆𝑅𝑖𝑡+𝑊𝑅 𝑗

𝑡+∆ 𝑡+𝑊𝑇 𝑗𝑡+∆𝑡+𝑊𝐿 𝑗

𝑡+∆𝑡−𝛿𝑖𝑗𝑡

Intersection iIntersection j

Equilibrium (NE)

Volume and Queue Length of Shared Lanes (VQ)

𝑟 𝑠 ,𝑎𝑡 =

𝑣𝑠𝑡

𝑣𝑎𝑡

𝑟 𝑟 ,𝑡𝑡 (𝑜𝑟 𝑟 𝑙 , 𝑡𝑡 )=¿¿

Equilibrium (NE)

Flow Characteristics of Shared Lanes (FC)

Equilibrium (NE)

• Not applicable in Nevada• Departure

detector

• Significant trip generator between intersections

• Network Equilibrium (NE)

• Downstream intersection• Volume and Queue Length of Shared Lanes

(VQ)• Not applicable if left turn is not protected

• Pedestrian

• Turning radius

•Flow Characteristics of Shared Lanes (FC)

• High variances in different hours, days, weeks, or seasons

•Observed proportions in the field (OP)

Considerations

Making the Models20

•User must specify the structure of the model

•Regression

•GP automatically evolves both the structure and the parameters

• Genetic Programmi

What does GP do?21

What does GP do?22

x y z3 5 5.8309528 14 16.12452

18 2 18.1107732 11 33.8378512 10 15.620521 6 21.840337 4 8.062258

16 24 28.844412 9 9.219545...

GP is based on evolution process, (cross over, mutation, and elitism)

We can use ready applications

How GP Works?23

GPTIPS24

MAPE: Mean Absolute Percentage Error

Di: the detector data valueBi: the reference (base) data valuen: the total number of intervals

Accuracy26

Case Study Intersections27

9th St. and Sierra St.

N McCarran Blvd and Clear Acre Ln

8th Street and Center Street

1 6 25 0.24 3 0.50

2 7 18 0.39 5 0.71

3 6 13 0.46 4 0.67

4 18 21 0.86 18 1.00

5 7 16 0.44 5 0.71

Sample of data collected for VQ method

1 0 0 1 c c T

1 2.24 2.24 2 c c R

1 2.31 4.55 3 t c T

1 3.01 5.32 4 c t T

Sample of data collected for FC method

MODELLING METHOD

N MCCARRAN BLVD/CLEAR ACRE LN 9TH ST/ SIERRA ST 8TH ST/ CENTER ST

Regression Not applicable Not applicable Not applicableGP Not applicable Not applicable Not applicable

Analytical Calculation

Not applicable due to un-controlled south bound

right turn

Not applicable due to downstream un-

signalized intersection

RegressionNot applicable because the adjoining lane is also

shared lane

Not applicable due to no existing same-direction adjoining lane

GPNot applicable due to no existing same-direction

adjoining lane

Not applicable due to no existing same-direction adjoining lane

Analytical Calculation

Not applicable Not applicable Not applicable

Regression * * *

GP * * *

Analytical Calculation Not applicable Not applicable Not applicable

Models30

MODELLING METHOD

INTERSECTION

TURNING RADIUS (FT)

TIME INTERVAL

MAPE (%)

Regression GPAnalytical

Calculation

FLOW CHARACTERISTI

CS (FC)

8th St and Center St

16Vehicle by

Vehicle20 14 -

Hourly Average 8 7 -

9th St and Sierra St

35Vehicle by

Vehicle21 20 -

McCarran Blvd and Clear Acre

Vehicle by Vehicle

28 25 -

VOLUME AND QUEUE (VQ)

8th St and Center St

Vehicle by Vehicle

17 15 -

NETWORK EQUILIBRIUM

9th St and Sierra St

Vehicle by Vehicle

Hourly Average - - 1

Results31

Results32

Vehicle by Vehicle

Hourly Average

Vehicle by Vehicle

Hourly Average

Vehicle by Vehicle

Hourly Average

Vehicle by Vehicle

Hourly Average

Vehicle by Vehicle

Hourly Average

8th St and Center St 9th St and Sierra St McCarran Blvd and Clear Acre Ln

8th St and Center St 9th St and Sierra St

Flow Characteristics (FC) Volume and Queue (VQ) Network Equilibrium (NE)

Mean Absolute Percentage Error (MAPE) percentRegression MAPE

GP MAPE

Analytical MAPE

Errors using inadequate data are much less than those using no data at

Charles Babbage in 1860

Considered “father of the computer”

Thank you

Any comments or questions?

If we have data, let’s look at data. If all we have are opinions, let’s go with mine! Jim Barksdale, former Netscape CEO

Ali Gholami Center for Advanced Transportation Education and Research (CATER)

researchuniversity of

renoproject panel meetingmarch

loop detector data

signalized intersections

loop detector5

detector information

turning movement proportions

intersection volume

Documents

Suzy Cater SYS Academy 2021

Men · LAMRI, Sid Ali B3 TURUMBETOV, Azamat B3 ARTAN,...

Scientific Research Bimonthly Vol.8, No.5 (Tome 40...

Sophie Cater Portfolio

Data Mining Instructor –Elham Gholami Contact...

Let Jonny Cater Your Needs - cdn.hibuwebsites.com

12.Sedighian & Gholami 1391, MAR 6 & 7

Cater-Pillars Catering - Virtual Enterprises...

2013 Cater y Recambio Tecnologico Empresas

DMP Residential Brochure - Cater Video & Security

2020 CATER - SpotHopper (static.spotapps.co)

Jaleh Gholami Eshlaghi MD. MPH. PhD Candidate in...

let us cater to you!

Tl Guide Hungry Cater Activity

We Love to Cater! We Deliver!

prace - cater