Simulate traffic on traffic on any section of Open Street Map! Inspired by A-B-street
Urban traffic simulator using OpenStreetMaps, Python, and PyGame.
- To-scale traveler speeds and map
- Uses gravity model to determine traveler generation and destination
- Flexible destination weighting
- Models congestion using road width and speed limit measurements, along with density of modeled travelers.
- Uses Dijkstra's shortest path algorithm for routefinding.
- Accounts for congestion to update traffic speeds and routefinding
Standard traffic sim window focused on a section of Brighton, MA. Street networks are outlined in black; travelers (people) in red; intersections in blue; attractions (destinations) in green.
Python3
Dijkstar
pip install Dijkstar
numpy
pip install numpy
Pygame
pip install pygame
cd src/- Adjust
main.pysuch that input map, screen width, and screen height are as desired python3 main.py
- Pygame window will open soon after (may take some time depending on size of map)
- Initial XML Received from OpenStreetMaps
- Parsed into readable and relevant JSON by xml_streets_to_json.py
- JSON includes dictionary of "nodes" (points in space)
- Nodes are split into two categories:
- "connections" are points along a road (connecting one location to another)
- "attractions" are other points on a map indicating places to be (residences, grocery stores, movie theaters)
- "ways" are collections of nodes, like a full road or a building
- Each way has a property of its "noderefs" the node ID of every node contained within it
- The following classes are defined by our program:
- Node: with property ID, type, longitude and latitude, and t
- Intersection: When a single node is a part of multiple ways, we can identify it as an intersection (a place where ways, AKA roads, meet)
- Properties noderef, list of outgoing way segments, list of incoming way segments
- Way segment: The parts of a road between intersections
- Each road has multiple way segments, split up by intersections
- This is done to allow travelers to move onto another road or continue on their current road at each intersection (Dijkstra's algorithm)
- Network: The collection of all intersections and way_segments that we observe in our data
- Traveler: Representation of cars travelling on roads from origin to destination nodes
- Engine: Runs the simulation
XML tags are parsed into JSON, into 3 dictionaries;
- Nodes - Points in space, define places or make up ways
- Subcategorized into "attractions" (more than just lon and lat) and connections (just lon and lat)
- Ways- Ordered lists of nodes which define a linear features (Ex: roads, train lines, rivers, building boundaries, ...)
- Subcategorized into "roads", "rails" and "other"
- Relations - Ordered List of Nodes, Ways and/or Relations, describes a special relation between Nodes (Ex: No Left Turn here)
- look more into: multipolygons
In each dictionary, individual elements are indexed by id. From there all of the properties of each element are indexed by descriptor. For "ways" and "nodes" a subcategory type is also needed to index before indexing by node:
- Ex:
my_attractions_node = json_dict["nodes"]["attractions"][NODE_ID_NUMBER]Key Descriptors: - Ways have a "name" tag and contain and ordered list of nodes Based on: https://wiki.openstreetmap.org/wiki/Elements
Note: dummy_streets.json doesn't have relations or ways for outlining buildings.
{
"nodes":{
"connections":{
"ID": {
"lat": "LATITUDE",
"lon": "LONGITUDE"
},
"ID": {
"lat": "LATITUDE",
"lon": "LONGITUDE"
},
...
},
"attractions":{
"ID":{
"lat": "LATITUDE",
"lon": "LONGITUDE",
"addr:city": "CITY",
"addr:housenumber": "HOUSE NUMBER",
"addr:postcode": "ZIP CODE",
"addr:state": "STATE",
"addr:street": "STREET ROAD",
...
},
"ID":{
"lat": "LATITUDE",
"lon": "LONGITUDE",
"addr:housenumber": "HOUSE NUMBER",
"addr:postcode": "ZIP CODE",
"addr:state": "STATE",
"addr:street": "STREET ROAD",
...
}
}
},
"ways": {
"roads" :{
"8814911": {
"node-refs": [
"NODE REF ID",
"NODE REF ID",
"NODE REF ID",
...
],
"condition": "fair",
"highway": "service",
"lanes": "1",
"massgis:way_id": "193636",
"source": "massgis_import_v0.1_20071009101959",
"width": "6.1"
},
},
"nonroads" : {
"ID": {
"node-refs": [
"NODE REF ID",
"NODE REF ID",
"NODE REF ID",
...
]
},
},
"rails":
"ID": {
"node-refs": [
"NODE REF ID",
"NODE REF ID",
"NODE REF ID",
...
],
"electrified": "ELECTRIFIED STATUS",
"frequency": "FREQUENCY",
"gauge": "GUAGE",
"massgis:geom_id": "GEOM ID",
"name": "NAME",
"railway": "RAIL TYPE",
"railway:preferred_direction": "FORWARD OR BACKWARD",
}
}
"relations": {
"9670662": {
"addr:housenumber": "HOUSE NUMBER",
"addr:street": "STREET NAME",
"building": "YES OR NO",
"type": "multipolygon",
"members": [
{
"type": "way",
"ref": "NODE REFERENCE",
"role": "outer"
},
{
"type": "way",
"ref": "NODE REFERENCE",
"role": "outer"
}
]
},
...
}
}