UAVFlightPath

Generates a boustrophedon (lawnmower) coverage flight path for a fixed-wing UAV over a rectangular survey area, with wind drift correction, terrain following, strip alignment, and Monte Carlo uncertainty analysis.

Quick start

pip install -r requirements.txt
python FlightPathGen.py

Outputs:

• flight_path.html — interactive Plotly figure (bird's-eye + altitude profile)
• flight_path.png — static PNG (kaleido; matplotlib fallback)
• monte_carlo.html — MC path spread + statistics dashboard
• monte_carlo.png — static PNG of MC results

Features

Feature	Description
Lawnmower path	East-West strips, alternating turn direction
Strip alignment	Repositioning leg after each turn corrects wind drift; strips start at exact nominal y positions
Spatially varying wind	WindField: base vector + 4-mode sine perturbations (smooth, realistic mesoscale variation)
Terrain following	TerrainModel: superposition of Gaussian hills; rate-limited climb/descent to maintain target AGL
Monte Carlo	run_monte_carlo: 100-run statistical sweep over airspeed, wind speed/direction, and nav noise
Coverage metrics	Strip deviation, coverage fraction, path length, reposition distance per run
Interactive plots	Plotly HTML for both nominal path and MC results; kaleido/matplotlib PNG export

Physics model

Quantity	Formula
Rate of turn	ROT (deg/s) = 1091 × tan(bank) / TAS(knots)
Ground speed	v_ground = v_airspeed_vector + v_wind_vector
G-load	n = 1 / cos(bank_angle)
Position	Euler integration at dt = 0.01 s
Terrain follow	z_target = terrain(x,y) + target_agl_ft, rate-limited by max_climb_rate_fpm

Rate of turn uses true airspeed, not ground speed. Ground speed is the vector sum of airspeed and wind (not subtraction).

Configuration

UAVConfig

Parameter	Default	Units	Description
map_x_km	1.0	km	Survey area east-west extent
map_y_km	1.0	km	Survey area north-south extent
airspeed_knots	20.0	knots	True airspeed (TAS)
max_g	2.0	G	Structural/comfort G-load limit — enforced at construction
bank_angle_deg	60.0	deg	Coordinated turn bank angle
wind_speed_mph	5.0	mph	Mean wind speed
wind_dir_deg	120.0	deg	Wind direction toward (from East, CCW)
swath_width_ft	100.0	ft	Camera footprint swath width — drives strip spacing
target_agl_ft	200.0	ft	Target altitude above ground level
max_climb_rate_fpm	500.0	ft/min	Maximum climb/descent rate for terrain following

UAVConfig.__init__ raises ValueError if bank_angle_deg produces G-load > max_g.

MonteCarloConfig

Parameter	Default	Description
n_runs	100	Number of simulation runs
airspeed_sigma_knots	1.0	Airspeed uncertainty 1-sigma (knots)
wind_speed_sigma_mph	2.0	Wind speed uncertainty 1-sigma (mph)
wind_dir_sigma_deg	15.0	Wind direction uncertainty 1-sigma (deg)
nav_sigma_ft	3.0	GPS/nav position noise 1-sigma per step (ft)

Known limitations

• 2-D strip alignment: repositioning leg snaps strip start to exact nominal y, but cross-wind drift during the leg still displaces the path relative to the strip centre line. Cross-track error correction (e.g. CTE guidance law) not yet modelled.
• Flat earth: coordinates are in feet, no map projection applied.
• Steady-state wind: WindField models smooth spatial variation but no temporal gusts.
• MC speed: each run uses the fast vectorised (uniform-wind) path generator for speed; per-run WindField terrain integration is available by passing wind_field= to generate_lawnmower_path at the cost of ~10× longer runtime per run.

Files

File	Purpose
FlightPathGen.py	Path generation, physics model, MC analysis, Plotly/matplotlib visualisation
Post.py	Post-processing stub (planned: GPS log alignment, coverage gap analysis, GeoJSON export)
requirements.txt	Python dependencies