3D Traffic Simulation Model

Based on a cellular automaton model 3D traffic simulation made with C++14.
3D engine made from scratch with OpenGL.

Building

App consists of 2 parts: simulation and visualization engine. The first one requires Boost to be installed (specifically property_tree). The latter (in MISS_Dispnew subdirectory) requires the same Boost library and additionally (Open)GL, GLEW, glm and X11 (xlib) libraries.

The app was tested with following library and tools versions:

• OS Ubuntu 20.04 and 22.04
• GCC 9 & 11,
• CMake 3.16 & 3.22
• Boost 1.71 & 1.74
• GLEW 2.1 & 2.2
• GLM 9.9.8

For convenience there's a docker builder container available (docker-builder directory). Assuming correctly configured docker, the build.sh script should download Ubuntu 20.04 image and build the application inside. The build.sh script expects 2 arguments - source directory and build directory. Example usage (from inside docker-builder subdirectory) that will create out directory in current directory:

./build.sh .. out

Usage

Run Debug/Sim It uses city model from file map.json and produces frames.json which is a set of frames used by 3D engine to display

Copy frames.json to MISS_Dispnew and run 3D engine cd MISS_Dispnew Debug/MISS_Disp

JSON city model

" nodes " : [ {
" id " : 0 ,
" lat " : -1 0 0 . 0 ,
" lon " : 0 . 1 ,
" intersection " : true ,
" connections " : [ {
" from " : " 0 r 0 " ,
" to " : [
"0r0,100"
]
} , {
" from " : " 0 r 1 " ,
" to " : [
"0r1,100"
]
} ],
" roads " : [
0, 1
],
} , ...

• node - a vertex = beginning or end of a road
• lon - longtitude, lat - latitude (in a simulated world, not real-world),
• intersection - is this object a crossroad [true/false]
• connections - list of roads that connects with such crossroad
• from - lane which from we driving,
• to - lane which to we can drive on such crossroad with probability of such action,
• 0r0,100 - probability of going to lane 0 on a road 0 is 100%,
• roads - list of roads which crossroad consist of.

" ways " : [ {
" id " : 0 ,
" beg " : 0 ,
" end " : 1 ,
" left " : 0 ,
" right " : 2 ,
" length " : 2 0 0 ,
" forbid " : [
{ " from " : " 0 l 0 " , " to " : " 0 l 1 " }
]
} ] , ...

• beg - node id which is beginning of a road,
• end - node id which is end of a road,
• left - number of left-side lanes on a road,
• right - number of right-side lanes on a road,
• length - length of a road,
• forbid - list of forbidden lane changes.

" lights " : [ {
" color " : " red " ,
" time " : 1 4 ,
" red " : 1 6 ,
" green " : 1 6 ,
" lanes " : [
"0l0", "0l1", "2l0", "2l1"
]
} , {
" color " : " green " ,
" time " : 1 4 ,
" red " : 1 6 ,
" green " : 1 6 ,
" lanes " : [
"1l0", "1l1", "3l0", "3l1"
]
} ] , ...

• color - light color at the start of the simulation [green/red],
• time - time left to light color to change,
• red - how long light is red,
• green - how long light is green,
• lanes - list of lanes which is controlled by such traffic light.

" spawner " : [ {
" lane " : "0r0",
" prob " : 15
} , {
" lane " : "0r1",
" prob " : 15
} , ...

It is an object that spawn new cars each update. There is said that on lane 0 on road 0 there is probability of 15% that a car will spawn.

Class Diagram