4. Weight for Road Network

The present model will introduce how to design a road system, or graph, based on the road GIS data and provide each edge a weight representing the destruction level of the road.

Formulation

• Add a destruction_coeff variable to the road agent. The value of this variable is higher or equal to 1 or lower or equal to 2. At initialization, the value of this variable is randomly defined between 1 and 2.
• In the road network graph, more a road is worn out (destruction_coeff high), more a people agent takes time to go all over it. Then the value of the arc representing the road in the graph is equal to "length of the road * destruction_coeff".
• The color of the road depends on the destruction_coeff. If "destruction_coeff = 1", the road is green, if "destruction_coeff = 2", the road is red.

Model Definition

road agent

We add a destruction_coeff variable which initial value is randomly defined between 1 and 2 and which have a max of 2. The color of the agent will depend on this variable. In order to simplify the GAML code, we define a new variable colorValue that represents the value of red color and that will be defined between 0 and 255.

species road  {
    float destruction_coeff <- rnd(1.0,2.0) max: 2.0;
    int colorValue <- int(255*(destruction_coeff - 1)) update: int(255*(destruction_coeff - 1));
    rgb color <- rgb(min([255, colorValue]),max ([0, 255 - colorValue]),0)  update: rgb(min([255, colorValue]),max ([0, 255 - colorValue]),0) ;
    ...
}

weighted road network

In GAMA, adding a weight for a graph is very simple, we use the with_weights operator with the graph for left-operand and a weight map for the right-operand. The weight map contains the weight of each edge: [edge1::weight1, edge2:: weight2,...]. In this model, the weight will be equal to the length of the road (perimeter of the polyline) its destruction coefficient.

    init {
        ...
        create road from: shape_file_roads ;
        map<road,float> weights_map <- road as_map (each:: (each.destruction_coeff * each.shape.perimeter));
        the_graph <- as_edge_graph(road) with_weights weights_map;
        ...
    }

Complete Model

model tutorial_gis_city_traffic

global {
    file shape_file_buildings <- file("../includes/building.shp");
    file shape_file_roads <- file("../includes/road.shp");
    file shape_file_bounds <- file("../includes/bounds.shp");
    geometry shape <- envelope(shape_file_bounds);
    float step <- 10 #mn;
    date starting_date <- date("2019-09-01-00-00-00");
    int nb_people <- 100;
    int min_work_start <- 6;
    int max_work_start <- 8;
    int min_work_end <- 16; 
    int max_work_end <- 20; 
    float min_speed <- 1.0 #km / #h;
    float max_speed <- 5.0 #km / #h; 
    graph the_graph;
	
    init {
	create building from: shape_file_buildings with: [type::string(read ("NATURE"))] {
	    if type="Industrial" {
		color <- #blue ;
	    }
	}
	create road from: shape_file_roads ;
	map<road,float> weights_map <- road as_map (each:: (each.destruction_coeff * each.shape.perimeter));
	the_graph <- as_edge_graph(road) with_weights weights_map;
		
	list<building> residential_buildings <- building where (each.type="Residential");
	list<building> industrial_buildings <- building  where (each.type="Industrial") ;
	create people number: nb_people {
	    speed <- rnd(min_speed, max_speed);
	    start_work <- rnd (min_work_start, max_work_start);
	    end_work <- rnd(min_work_end, max_work_end);
	    living_place <- one_of(residential_buildings) ;
	    working_place <- one_of(industrial_buildings) ;
	    objective <- "resting";
	    location <- any_location_in (living_place); 
	}
    }
}


species building {
    string type; 
    rgb color <- #gray  ;
	
    aspect base {
	draw shape color: color ;
    }
}

species road  {
    float destruction_coeff <- rnd(1.0,2.0) max: 2.0;
    int colorValue <- int(255*(destruction_coeff - 1)) update: int(255*(destruction_coeff - 1));
    rgb color <- rgb(min([255, colorValue]),max ([0, 255 - colorValue]),0)  update: rgb(min([255, colorValue]),max ([0, 255 - colorValue]),0) ;
	
    aspect base {
	draw shape color: color ;
    }
}

species people skills:[moving] {
    rgb color <- #yellow ;
    building living_place <- nil ;
    building working_place <- nil ;
    int start_work ;
    int end_work  ;
    string objective ; 
    point the_target <- nil ;
		
    reflex time_to_work when: current_date.hour = start_work and objective = "resting"{
	objective <- "working" ;
	the_target <- any_location_in (working_place);
    }
		
    reflex time_to_go_home when: current_date.hour = end_work and objective = "working"{
	objective <- "resting" ;
	the_target <- any_location_in (living_place); 
    } 
	 
    reflex move when: the_target != nil {
	do goto target: the_target on: the_graph ; 
	if the_target = location {
	    the_target <- nil ;
	}
    }
	
    aspect base {
	draw circle(10) color: color border: #black;
    }
}


experiment road_traffic type: gui {
    parameter "Shapefile for the buildings:" var: shape_file_buildings category: "GIS" ;
    parameter "Shapefile for the roads:" var: shape_file_roads category: "GIS" ;
    parameter "Shapefile for the bounds:" var: shape_file_bounds category: "GIS" ;	
    parameter "Number of people agents" var: nb_people category: "People" ;
    parameter "Earliest hour to start work" var: min_work_start category: "People" min: 2 max: 8;
    parameter "Latest hour to start work" var: max_work_start category: "People" min: 8 max: 12;
    parameter "Earliest hour to end work" var: min_work_end category: "People" min: 12 max: 16;
    parameter "Latest hour to end work" var: max_work_end category: "People" min: 16 max: 23;
    parameter "minimal speed" var: min_speed category: "People" min: 0.1 #km/#h ;
    parameter "maximal speed" var: max_speed category: "People" max: 10 #km/#h;
	
    output {
	display city_display type: opengl {
	    species building aspect: base ;
	    species road aspect: base ;
	    species people aspect: base ;
	}
    }
}