al-wang / ocean

OCEAN for Multi-Resolution Network Modeling:

Open-source tool for creating Cell-based (microscopic) and mEsoscopic and mAcro Networks

Development Support:

Ph.D. Student: Jiawei (Jay) Lu ([email protected])

Dr. Xuesong (Simon) Zhou ([email protected])

School of Sustainable Engineering and the Built Environment

Arizona State University

Overview of work flow

Step 1: Obtain macroscopic network data

Use the OSM2GMNS tool to convert map.osm file in OSM format into a network file in GMNS format, and the output includes node.csv and link.csv files. The OSM2GMNS tool also generates important information such as the mapping between “model nodes” and “main signal nodes”, which is recorded in file complex_intersection.csv and segment.csv.

As shown above, ensure the following files are located in the “consolidated” folder:

node.csv (required),

link.csv(required),

complex_intersection.csv(optional),

segment.csv(optional).

Step 2: Obtain the OCEAN python code

Go to the github website of OCEAN,

https://github.com/asu-trans-ai-lab/Ocean/tree/master/ocean

download the source codes insides the “ocean” folder, and open main.py file.

Step 3: Configure network-related parameters

Check and configure parameter values in the python code of main.py.

Descriptions of parameters in the code

working_directory	File directory of node.csv and link.csv
coordinate_type	Type of coordinate system, m: meter, ll: latlon, f: feet
geometry_source	File source of geometry information, n: none, l: link file, g: geometry file
unit_of_length	Minimum unit of link length, m: meter, km:kilometer, mi: mile, f: feet
segment_unit	length unit for segment data, m: meter, km:kilometer, mi: mile, f: feet
speed_unit	unit of traveling speed, mph: mile per hour, kph: kilometer per hour
link_type_list	List of allowed link types
connector_type	link type of connectors in macroscopic network, -1 if no connector
The following parameters have default values
min_link_length	meter, links shorter than that will be removed, > 2 * length_of_cut[0]
comb_links	remove 2-degree nodes
auto_connection	generate movement connections for intersections without predefined movement info
connector_geometry_for_output	1: with lane offset; 2: no lane offset
length_of_cell	Length of a cell when generating microscopic networks from macro network
length_of_cut	Information of offset, e.g, 2:8.0 cut 8 meters if the original macro link has 2 lanes, etc
cells_in_queue_area	for signalized intersections
width_of_lane	Width of each lane

Step 4: Execute OCEAN and generate networks

Go to the github website of OCEAN, https://github.com/asu-trans-ai-lab/Ocean/tree/master/ocean/test/asu_dataset**,** download a test data set or use your own dataset.

Run the OCEAN program to generate three sets of file corresponding to macroscopic, mesoscopic, and microscopic network in the folders shown below. Each folder should have at least node.csv and link.csv files.

Step 5: Visualize networks in NeXTA

You can open and visualize the traffic network project (node.csv in three folds) in NEXTA .

Use menu as shown below, then set the window layout to show three levels of networks, through menu items “tile vertically” and “Synchronized Display” .

The potential next step is to manage OD zone structure and in the NeXTA tool and perform traffic assignment and simulation using DTALite for transportation network simulation and analysis.