Skip to content

🧭 Trips Generator 📦

Status Badge Module Type Internal Usage

This module converts Sevilla traffic counts into SUMO trips and routes (.rou.xml) using a configurable mapping and time-of-day distributions. It complements the Traffic Profile Analysis by turning demand estimates into simulation-ready inputs.

Author: Code — Raquel Blanco · Docs — Belén López Salamanca
Date: 07/2025
Version: 1.0.0

📝 Project Overview

The trips generator reads cleaned interval counts for Sevilla routes, as prepared in Traffic Data. Route identifiers are manually mapped to SUMO edge IDs in rutas_edges_v1.json. With those inputs, the Python script (randomTrips_v7) produces a SUMO route file (.rou.xml) that follows the observed temporal distribution of vehicles.

🧰 Inputs

CSV with route vehicle counts

  • Path: In Sanevec-docs:
    traffic_profiling/trips/datos_combinados_1.csv
  • Description: See Traffic Data.
  • Expected columns:
  • RUTAS: integer route identifiers (e.g., 1, 2, 3, …).
  • Hour columns: per-route counts for each time window (e.g., 07:00-08:00, 08:00-09:00, …).
  • Optional Total column: may be present; ignored.

JSON with route edges

  • Path: In Sanevec-docs:
    traffic_profiling/trips/rutas_edges_v1.json
  • Description: Manually curated mapping that links each route (RUTA_n) to one or more SUMO edge IDs from the Sevilla network.
  • Structure: Values are comma-separated SUMO edge IDs used as candidate source edges for each route. 
    {
  "RUTA_1": "edgeA, edgeB, edgeC",
  "RUTA_2": "edgeX, edgeY"
}

SUMO network

  • Path: osm.net.xml
  • Description: Map of Seville obtained with the OSM Web Wizard.
  • Usage: Compute shortest paths and filter target edges (allows("passenger") and not edge.is_fringe()).

🧪 Environment

  • SUMO_HOME set so sumolib tools are available.
  • MPI environment available for mpi4py (hours distributed across ranks).

📤 Outputs

  • Route file for SUMO: generated_routes.rou.xml
    Contains <vehicle id="..." depart="..."><route edges="..."/></vehicle> entries, sorted by depart time.
  • Console logs: Per-route/hour generation stats and final totals.

🧼 Pre-processing

  1. Load vehicle counts from CSV. Treat all columns except RUTAS and Total as hour bins.
  2. Load edges mapping from JSON. For each RUTA_n, split by comma and trim whitespace.
  3. Build per-hour objects:
  4. For each hour, create Route(route_id=f"{RUTA}_{hour}", total_vehicles=n, edges=[...]).
  5. Store (routes_for_hour, total_vehicles_in_hour) under the hour key.
  6. Read the SUMO network and filter target edges: non-fringe and passenger-allowed.

⚙️ Generation Logic

Per MPI rank: 1. Distribute hours across ranks (even split; remainder on lower ranks). 2. For each local hour:

  • Uniform depart spacing across the hour: 3600 / total_vehicles_in_hour.

  • For each Route:

    • Pre-assign overlapped vehicles (shared edges with previously generated routes in the same hour) to reduce new generation.
    • While the route still needs vehicles:
    • Sample a depart time, a source edge from the route, and a target edge from global valid targets.
    • Compute shortest path (source → target). If none, retry.
    • Create Vehicle(id, depart_time, path_edges).
    • Penalize the route’s edges (inflate first-lane length) to discourage re-use in subsequent shortest-path calls (heuristic).
    • Record overlap so remaining routes can pre-assign accordingly.
    • Gather all vehicles at rank 0 and write the .rou.xml file.

📚 Vehicle Route Generation — Docstrings Extracted

Module

Vehicle route generation module for SUMO traffic simulations. 
This module generates vehicles routes based on input data

Classes

Vehicle
Class Docstring
Represents a vehicle with its route an departure time
Vehicle.__init__(id, depart_time, edges)
Initialize a Vehicle instance

Args:
    route_id: unique identifier for the vehicle
    depart_time: Simulation time when the vehicle departs (seconds)
    edges: List of edge IDs that form the vehicle's route
Vehicle.__repr__()
Return a string representation of the Vehicle.
Route

Class docstring: None provided.

Route.__init__(route_id, total_vehicles, edges)
Initialize a Route instance
Args:
    route_id: unique identifier for the route
    total_vehicles: number of vehicles assigned to this route
    edges: list of edges ids that form the route
Route.add_vehicles(id_vh_list)
Add vehicles to the processed vehicles list.

Args:
    vehicles: List of Vehicle objects to add
Route.processed()

Docstring: None provided.

Route.__repr__()

Docstring: None provided.

Functions

load_route_data(route_vehicles_file_path, route_edges_file_path) -> List[Route]
Load route data from CSV and JSON files.

Args:
    route_vehicles_file_path: Path to CSV file with vehicle counts per route and hour
    route_edges_file_path: Path to JSON file with edge sequences for each route

Returns:
    Dictionary mapping hours to tuples of (routes, total_vehicles)
find_routes_by_edge(routes, current_id, vehicles, update_routes: dict)
Find routes that share edges with the given vehicles.

Args:
    routes: List of all Route objects
    current_id: ID of the current route being processed
    vehicles: List of Vehicle objects to check
    update_routes: Dictionary to accumulate route updates

Returns:
    Updated dictionary mapping route IDs to lists of vehicle IDs
generate_vehicles_for_hour(start_time, net: Net, target_edges, routes: List[Route], total_vh, start_id: int) -> List[Vehicle]
Generate vehicles for a specific hour.

Args:
    start_time: Starting time for this hour (seconds)
    net: SUMO network object
    target_edges: List of valid target edges for vehicles
    routes: List of Route objects for this hour
    total_vh: Total number of vehicles to generate for this hour
    start_id: Starting ID for vehicle numbering

Returns:
    Tuple of (list of generated vehicles, next available vehicle ID)
create_route_file(vehicles: List[Vehicle], filename="generated_routes.rou.xml")
Create a SUMO route file from a list of Vehicle objects.

Args:
    vehicles: List of Vehicle objects to include in the file
    filename: Name of the output file

🔮 Next Steps

  • Integrate directionality and multi-class vehicle types.
  • Add OD matrices or downstream edge selection for longer routes.
  • Validate against --edgedata-output metrics and refine distributions.

📎 Notes & References