Drone Simulation with PID Flight Controller

Advanced drone simulation featuring realistic flight dynamics, collision detection, and real-time telemetry visualization.

Features

🚁 Realistic Flight Controller

• Cascaded PID control loops (rate → attitude → position)
• Multiple flight modes: Stabilize, Altitude Hold, Position Hold
• Conservative gains to maintain horizontal flight attitude

🎯 Advanced Controls

• Real-time sliders for target positioning
• Direct coordinate input fields
• Position hold buttons (current location lock)
• One-click simulation reset

💥 Physics-Based Collision Detection

• Real MuJoCo collision detection with obstacle geometry
• Crash behavior with PID disable and gravity tumbling
• Collision depth measurement and logging
• Emergency disarm on severe impacts

📊 Comprehensive Telemetry

• Real-time position display (X, Y, Z coordinates)
• Speed monitoring (individual axes + total speed)
• Distance to target calculation
• Attitude angle limits to prevent excessive tilt

🏗️ 3D Environment

• Multiple obstacle types: towers, boxes, spheres, pillars
• Room boundaries with walls and ceiling
• Visual and physics collision matching

Files

Core Simulation (3 files only)

• drone_collision_course.py - Main simulation with all features
• flight_controller.py - PID-based flight controller implementation
• drone_with_obstacles.xml - MuJoCo world with physics obstacles

Installation

1. Create virtual environment:

python3 -m venv venv
source venv/bin/activate

2. Install dependencies:

pip install -r requirements.txt

Usage

1. Run Simulation:

source venv/bin/activate
python drone_collision_course.py

2. Open Browser: Navigate to http://localhost:8082

3. Flight Operations:

   • Click "ARM" to enable the flight controller
   • Use "POSITION HOLD" or "🔒 HOLD CURRENT POSITION" to start flying
   • Adjust target position with sliders or direct coordinate input
   • Monitor real-time telemetry in the GUI

Flight Controller Details

PID Architecture

• Rate Controllers: Roll/Pitch/Yaw angular velocity control
• Attitude Controllers: Roll/Pitch angle stabilization
• Position Controllers: X/Y position hold with tilt angle limits
• Altitude Controller: Z-axis position control with thrust adjustment

Key Improvements

• Conservative Gains: Reduced from kp=1.0 to kp=0.2 to prevent overshoot
• Attitude Limits: Maximum tilt angle limited to ~8.5° (was 28.6°)
• Velocity Damping: Speed-based gain reduction when moving >1 m/s
• Collision Recovery: PID disable during crashes with physics takeover

Controls

Control	Description
ARM/DISARM	Enable/disable flight controller
POSITION HOLD	Fly to slider target coordinates
🔒 HOLD CURRENT	Lock drone at current position
🔄 RESET	Reset simulation to initial state
Target Sliders	Adjust X, Y, Z target positions
Direct Input	Enter precise coordinates

Telemetry Display

• Current Position: Real-time X, Y, Z coordinates
• Speed: Individual axis velocities + total speed
• Distance to Target: 3D and horizontal distance
• Collision Counter: Number of detected crashes
• Last Crash: Most recent collision details

Collision Detection

The simulation features realistic collision detection:

• Physics-based contact detection using MuJoCo
• Collision depth measurement
• Automatic PID disable on impact
• Emergency disarm for severe crashes (>5cm penetration)
• Random tumbling motion during crashes

Development

Built with:

• MuJoCo - Physics simulation and collision detection
• Viser - Real-time 3D visualization and GUI
• NumPy - Mathematical operations and array handling

Troubleshooting

Drone won't respond:

• Ensure drone is ARMED first
• Check that a flight mode is selected (POSITION HOLD)
• Verify target coordinates are within room boundaries

Excessive oscillation:

• PID gains are already conservative - check for crashes
• Use RESET to clear any stuck states
• Position hold may take time to stabilize

Performance issues:

• Close other browser tabs using port 8082
• Restart simulation if memory usage is high

---

Generated with Claude Code and optimized for realistic drone flight simulation.