ASPaeroFlow Data Generator

ASPaeroFlow Data Generator is a Python-based toolkit to create realistic air traffic scenarios for research and simulations. It builds synthetic flight schedules and airspace structures from historical data (e.g. OpenSky Network), and prepares inputs for strategic Air Traffic Flow and Capacity Management (ATFCM) analyses. In the overall ASPaeroFlow pipeline, this data generator is responsible for producing the traffic scenarios – including flights, navigation points, and sector definitions – that can be fed into optimization models or simulation tools.

Features

• Synthetic Traffic Scenario Generation: Construct a full day of flights based on statistical models derived from real data (e.g. OpenSky flight records).
• OpenSky Data Integration: Build probabilistic models of departures, destinations, durations, and turnarounds from OpenSky flight listings.
• Navpoint Graph via Voronoi Partitioning: Generate a navigation graph of waypoints and airports using geometric criteria (e.g., RNG or Gabriel Graph).
• Sector Generation and Capacities: Partition airspace into sectors with configurable size and assign realistic capacities to each.
• Configurable Parameters: Full control over flights, traffic scaling, navgraph shape, regions, and temporal resolution.
• BlueSky Simulation Compatibility: Convert generated scenarios to .scn files for playback in BlueSky ATC simulator.
• Optimizer-Ready Output: Output flights, sectors, and navgraph in CSV format for downstream optimization models.

Installation

Recommended platform: Linux (tested). Basic compatibility exists for macOS and Windows (see below).

1. Clone the Repository:

   git clone https://github.com/YourUsername/ASPaeroFlow-DataGenerator.git
   cd ASPaeroFlow-DataGenerator

2. Install Dependencies:

   pip install numpy pandas networkx tqdm

3. Download Required Data:

   • OpenSky COVID19 Flight Dataset (cc-by) → Place e.g., flightlist_20190601_20190630.csv in ./data/.
   • OurAirports airports.csv → Place in ./ourairports/.
   • Optionally: earth_fix.dat and earth_nav.dat from X-Plane → Place in ./test_navpoints/.

4. Platform Notes:

   • macOS: Should work if Python and dependencies installed.
   • Windows: Use WSL or adjust paths and shell calls accordingly.

Usage

Run the full scenario generator:

python run_pipeline.py \
  --csv-path data/flightlist_20190601_20190630.csv \
  --target-day 2019-06-15 \
  --config configs/dach_region.json \
  --experiment-name DACH-2019-06-15 \
  --scale 0.5 \
  --seed 42

Common Options

Option	Description
--csv-path	Path to OpenSky CSV (or folder of them)
--target-day	Day to simulate (YYYY-MM-DD)
--scale or --flights	Scale traffic by factor or use exact flight count
--experiment-name	Folder name under unparsed_experiment_data/
--grid-navpoints true	Use synthetic navpoint grid instead of real fixes
--sector-default-navaid-size N	Approx. navpoints per sector (for grouping)
--time-granularity	Time slot size (e.g. 4 = 15 min)
--seed	Random seed for reproducibility

See --help for more.

Outputs

Generated in unparsed_experiment_data/<experiment-name>/

• model/: Statistical models (departure rates, durations, etc.)
• navgraph/:
  • vertices.csv: Waypoints and airports
  • edges.csv: Navgraph edges
  • sectors.csv: Sector capacities
  • navaid_sector_assignment.csv: Sector grouping (if used)
• DATA_S*/:
  • flights.csv, aircrafts.csv
  • filed_flights.csv: Flight plans (waypoints + timestamps)
  • manifest.json: Summary of the run

Optional:

• experiment_data/: Transformed version for solvers
• bluesky/: .scn file for BlueSky simulator

BlueSky Conversion

python 06_bluesky_converter.py \
  --data-dir unparsed_experiment_data/DACH-2019-06-15/DATA_S0p5_42 \
  --navgraph-dir unparsed_experiment_data/DACH-2019-06-15/navgraph \
  --time-granularity 60

Repository Structure

Script	Purpose
00_model_generation_script_refactored.py	Learns statistical traffic model
01_data_generation_script_refactored.py	Samples synthetic flights
02_graph_generator.py	Builds navgraph (grid or real fixes)
03_sector_capacity_generator.py	Assigns sector IDs and capacities
04_simplified_filed_flight_plan_generator.py	Generates routed flight plans
05_transform_for_optimizer.py	(Optional) Converts for solver input
06_bluesky_converter.py	(Optional) Creates BlueSky .scn file
07_check_parsed_experiments_graph_connectedness.py	(Debugging) Checks graph connectivity

Development

• Custom Regions: Use --config JSON with a polygon to restrict to a geographic area.
• Navgraph Tweaks: Adjust RNG vs Gabriel Graph, or grid shape and spacing.
• Sectorization Control: Use BFS grouping or convex grouping mode.
• Multiple Scenarios: Use --flight-scales and --flight-seeds as lists.
• Testing: Run small scenarios first and validate route feasibility.

License

MIT license with attribution (for details see license.md).

---

This README describes the ASPaeroFlow-DataGenerator for air traffic research. Outputs are compatible with BlueSky, optimization models, or ASP-based planning. This README was created with the help of generative AI.