This python tool aims to offer a light-weight computational engine to generate optimize signal control timing data, and analyze the effectiveness of signal control strategies. Our goal is to automate the process of optimizing movement-based, phase-based signal control strategy and provide the interfaces for other mesoscopic and microscopic Analysis, Modeling and Simulation (AMS), based on Highway Capacity Manual’s Quick Estimation Method, and GMNS based network files.
The users and students can follow the steps below to learn and use this set of tools.
Demonstration of the tool w/ Video: https://youtu.be/8YB5WF4R550 (Contributed by Sohum Berdia @ASU)
Step 1: Familiar with Excel based Tool
Please check out the tool developed by Dr. Milan Zlatkovic at University of Wyoming (http://www.uwyo.edu/civil/faculty_staff/faculty/milan-zlatkovic/index.html, [email protected]) for his Excel based Quick Estimation Method
1.1. Excel files: milan1981/Sigma-X: Excel-based computational engine for signalized intersections (github.com)
1.2: Youtube teaching videos:
-
QGIS visualization video: QGIS Visualization for GMNS Traffic Signal Data – (contributed by Fan Yu @ASU)
https://www.youtube.com/watch?v=6hoYJtEaTn4&t=8s
-
Introduction to Synchro 11: https://www.youtube.com/watch?v=XXGOJl_7Owk&t=60s (contributed by Sohum Berdia @ASU)
-
Introduction of UTDF2GMNS (contributed by Fan Yu @ASU)
https://www.youtube.com/watch?v=-m_szNHdWoo&t=1s
1.3: Research paper: Zlatkovic, Milan, and Xuesong Zhou. "Integration of signal timing estimation model and dynamic traffic assignment in feedback loops: System design and case study." Journal of Advanced Transportation 49, no. 6 (2015): 683-699. (Open Access)
1.4. Try to test different results from different movement configurations at https://github.com/asu-trans-ai-lab/vol2timing/tree/master/examples/ 3_simple_cases
Case 3: Major EW 3 lanes
CL=50 s
Case 4L: Major EW 4 lanes with left turn volume
CL =150 s
Case 4: Major EW 4 lanes without left turn volume
CL =80 s
#Highlights:
Planning-level analysis of existing intersections
Estimation of signal timing parameters for known inputs
Major elements: Left-turn treatment; Lane volume; Signal timing; Critical intersection volume-to-capacity ratio; Control delay & LOS
Step 2: Python package-based tool
This python version aims to generate movement-based, phase-based signal control strategy, developed by the research teams led by Dr. Xuesong (Simon) Zhou at Arizona State University ([email protected]) and Dr. Milan Zlatkovic at University of Wyoming
Table 1. Folders of Vol2Timing package
| Github Folder Name | Contents |
|---|---|
| src | source code of Vol2Timing |
| Release | Python test code and data set |
| Doc | User’s guide and other documentations for Vol2Timing |
| Examples | Simple test cases and real world examples |
2.1. Test data set at for ASU network
https://github.com/asu-trans-ai-lab/vol2timing/tree/master/release
In this release folder, we have input files of (1) node.csv, (2) link.csv (3) movement.csv, and output files of (1) signal_timing_phase.csv, (2) signal_phase_mvmt.csv in GMNS format, as well as (3) timing.csv for quick mesoscopic and microscopic simulation.
To visualize the location of signal intersections, please create a node.csv contains only nodes with field osm_highway== “traffic_signals”, and use https://asu-trans-ai-lab.github.io/index.html#/ to visualize the locations of signalized intersections.
2.2. Try test script at
(1) Python code at https://github.com/asu-trans-ai-lab/vol2timing/blob/master/release/test.py
(2) Jupyter Notebook at https://github.com/asu-trans-ai-lab/vol2timing/blob/master/vol2timing.ipynb
(3) Utah State Street example from the original osm file at
https://github.com/asu-trans-ai-lab/vol2timing/tree/master/examples/Utah_state_street
2.3. For advanced users, you can check out our source code at https://github.com/asu-trans-ai-lab/vol2timing/tree/master/src, and learn the steps from the log file (*.log)
Other useful references:
(1) Nourmohammadi, Fatemeh, Mohammadhadi Mansourianfar, Sajjad Shafiei, Ziyuan Gu, and Meead Saberi. "An Open GMNS Dataset of a Dynamic Multi-modal Transportation Network Model of Melbourne, Australia." Data 6, no. 2 (2021): 21.
https://www.mdpi.com/2306-5729/6/2/21/htm
Pleas check out the nice relational structure of the developed General Modeling Network Specification (GMNS) dataset, in Fig. 1.
Step 3: Practice problem
Fig.1 Sample problem from text book
Step 1. Input information
First, we should extract movement, volume, and lane information from Fig.1.
| Movement | EBL | EBT | EBR | WBL | WBT | WBR | NBL | NBT | NBR | SBL | SBT | SBR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Volume | 300 | 900 | 200 | 250 | 1000 | 150 | 90 | 340 | 50 | 70 | 310 | 60 |
| No. of lanes | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
| Shared lanes | - | - | - | - | - | - | - | - | - | - | - | - |
Step 2. Select signal phasing
2.1 Determine left-turn treatment
There are several principles for determining left-turn treatment.
- Left-turn lane check
Criterion: If the number of left-turn lane on any approach exceeds 1, then it is recommended that the left turns on that approach be protected.
- Minimum volume Check
Criterion: If left-turn volume on any approach exceeds 240 veh/h, then it is recommended that the left turns on that approach be protected.
- Opposing Through Lanes Check
Criterion: If there are more than 4 or more through lanes on the opposing approach, then it is recommended that the left turns on that approach be protected.
- Opposing Traffic Speed Check
Criterion: If the opposing traffic speed exceeds 45mph, then it is recommended that the left turns on that approach be protected.
- Minimum Cross-Product Check
Criterion:
Protected+permissive:
| Number of Through Lanes | Minimum Cross-Product |
|---|---|
| 1 | 50000 |
| 2 or more | 100000 |
Protected only:
| Number of Through Lanes | Minimum Cross-Product |
|---|---|
| 1 | 150000 |
| 2 or more | 300000 |
Calculation: cross-product for each left-turn
| Movement | EBL | WBL | NBL | SBL |
|---|---|---|---|---|
| Opposing Through Lanes | ||||
| Cross-Product | ||||
| Exceed Protected Minimum Cross-Product?(Y/N) | ||||
| Exceed Protected+Permissive Minimum Cross-Product?(Y/N) | ||||
| Protected decision |
- Comparing results
Based on the analysis above, we can reach the left-turn final decision.
QEM result
| Movement | EBL | WBL | NBL | SBL |
|---|---|---|---|---|
| Left-turn Treatment | Protected+Permissive | Protected+Permissive | Permitted | Permitted |
Vol2Timing result
| Movement | EBL | WBL | NBL | SBL |
|---|---|---|---|---|
| Left-turn Treatment | Protected | Protected | Protected | Protected |
- Determine phasing sequence
QEM result
| Final Ring-Barrier Movement | ||||
|---|---|---|---|---|
| Ring1 | 1 | 2 | 3 | 4 |
| Ring2 | 5 | 6 | 7 | 8 |
Vol2Timing result
| Final Ring-Barrier Movement | ||||
|---|---|---|---|---|
| Ring1 | 1 | 2 | 4 | |
| Ring2 | 5 | 6 | 8 |
Step 3. Compute flow rates & adjusted saturation flow rate
Right-turn Movement
Left-turn Movement
Through Movement
Through Movement with exclusive LT lane & shared LT lane
Saturation flow rate
for protected phase:
The default value of saturation flow rate for protected phase is 1530veh/h/lane
for permissive phase(for left-turn):
The default value of saturation flow rate for permissive phase is 150-200veh/h/lane
Comparing results
Based on the analysis above, we can calculate saturation flow rate.
QEM result
| Movement | EBL | EBT | EBR | WBL | WBT | WBR | NBL | NBT | NBR | SBL | SBT | SBR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Saturaion flow rate | 1530 | 1530 | 1530 | 1530 | 1530 | 1530 | 0 | 1530 | 1530 | 0 | 1530 | 1530 |
Vol2Timing result
| Movement | EBL | EBT | EBR | WBL | WBT | WBR | NBL | NBT | NBR | SBL | SBT | SBR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Saturaion flow rate | 1530 | 1530 | 1530 | 1530 | 1530 | 1530 | 1530 | 1530 | 1530 | 1530 | 1530 | 1530 |
Step 4. Determine critical lane groups & total cycle lost time
（1）Determine critical lane groups
To determine the critical lane group for each stage, we should select the lane group with maximum v/s(v: volume, s: saturation rate) for each stage.
QEM result
| Stage | 1 | 2 | 3 |
|---|---|---|---|
| Critical lane group | EBL | WBT | NBL |
Vol2Timing result
| Stage | 1 | 2 | 3 | 4 |
|---|---|---|---|---|
| Critical lane group | EBL | WBT | NBL | NBT |
（2）Determine total cycle lost time
The total cycle lost time is given as
where
QEM result
= 4s*3 = 12s
Vol2Timing result
= 4s*4 = 16s
Step 5. Calculate cycle length
A practical equation for the calculation of the cycle length that seeks to minimize
vehicle delay was developed by Webster (1969). Webster’s optimum cycle length
formula is
where
QEM result
Vol2Timing result
Step 6. Allocate green time
There are several strategies for allocating the green time to the various stages. One of the most popular and simplest is to distribute the green time so that the v/c ratios are equalized for the critical lane groups, as by the following equation:
where
QEM result
| Stage | 1 | 2 | 3 | 4 |
|---|---|---|---|---|
| Ring1 | 24 | 86 | 0 | 30 |
| Ring2 | 28 | 82 | 0 | 30 |
Vol2Timing result
| Stage | 1 | 2 | 3 | 4 |
|---|---|---|---|---|
| Ring1 | 9 | 36 | 5 | 12 |
| Ring2 | 9 | 36 | 5 | 12 |
Step 7. Calculate capacity and V/C ratio
Capacity can be calculated as follows:
where
Then we can calculate the ratio of V/C.
QEM results
| Movement | EBL | EBT | EBR | WBL | WBT | WBR | NBL | NBT | NBR | SBL | SBT | SBR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Capacity(veh) | 338 | 896 | 896 | 294 | 852 | 852 | 65 | 284 | 284 | 65 | 284 | 284 |
Vol2Timing results
| Movement | EBL | EBT | EBR | WBL | WBT | WBR | NBL | NBT | NBR | SBL | SBT | SBR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Saturaion flow rate | 280.5 | 969 | 969 | 280.5 | 969 | 969 | 178.5 | 357 | 357 | 178.5 | 357 | 357 |
Step 8. Calculate Signal Delay and LOS
QEM results
| Movement | EBL | EBT | EBR | WBL | WBT | WBR | NBL | NBT | NBR | SBL | SBT | SBR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Phase | 5 | 2 | 2 | 1 | 6 | 6 | 0 | 8 | 8 | 0 | 4 | 4 |
| UnifDelay (s) | 16.1 | 29 | 13.62 | 19.06 | 31 | 15.06 | 57 | 57 | 47.82 | 57 | 57 | 48.11 |
| IncDelay (s) | 27.65 | 30.07 | 0.5 | 25.25 | 89 | 0.4 | 242.05 | 118.81 | 1.19 | 135.4 | 79.57 | 1.48 |
| ControlDelay (s) | 43.8 | 59.1 | 14.1 | 44.3 | 120.0 | 15.5 | 299.1 | 175.8 | 49.0 | 192.4 | 136.6 | 49.6 |
| LOS | D | E | B | D | F | B | F | F | D | F | F | D |
| Approach Delay (s) | 49.4 | 95.3 | 185.7 | 133.6 | ||||||||
| Approach LOS | D | F | F | F | ||||||||
| Intersection Delay (s) | 95.1 | |||||||||||
| Intersection LOS | F | |||||||||||
| Intersection V/C | 1.07 | |||||||||||
| Intersection Status | Over Capacity |
Vol2Timing results
| Movement | EBL | EBT | EBR | WBL | WBT | WBR | NBL | NBT | NBR | SBL | SBT | SBR |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Phase | 5 | 2 | 2 | 1 | 6 | 6 | 0 | 8 | 8 | 0 | 4 | 4 |
| UnifDelay (s) | 433.7 | 196.0 | 136.5 | 270.0 | -40.3 | 107.2 | 91.1 | 944.3 | 120.7 | 85.2 | 553.5 | 82.6 |
| LOS | D | E | B | D | F | B | F | F | D | F | F | D |
| Approach Delay (s) | 95.8 | 42.8 | 640.1 | 380 | ||||||||
| Approach LOS | F | D | F | F | ||||||||
| Intersection Delay (s) | 95.1 | |||||||||||
| Intersection LOS | F | |||||||||||
| Intersection V/C | 1.07 | |||||||||||
| Intersection Status | Over Capacity |