Auto Simulation System

A comprehensive autonomous vehicle simulation system with advanced perception, decision-making, and control capabilities.

Features

Core Functionality

• Traffic Light System: Set and control traffic lights with left/right/straight signals
• Obstacle Management: Place various types of obstacles (cones, barriers, etc.)
• Pedestrian Simulation: Realistic pedestrian behavior with trajectory prediction
• Vehicle Control: Advanced control algorithms for autonomous driving

Sensor Systems

• Camera: Visual perception for object detection
• LiDAR: 3D point cloud sensing for precise distance measurement
• Radar: Robust detection in adverse weather conditions
• Multi-sensor Fusion: Kalman filter-based sensor fusion for accurate object tracking

Advanced Algorithms

• SLAM: Simultaneous Localization and Mapping with particle filters
• BEV System: Bird's Eye View visualization for situational awareness
• Trajectory Prediction: Advanced pedestrian and vehicle trajectory prediction
• Collision Avoidance: Multi-lane pedestrian and obstacle avoidance

Control Algorithms

• PID Control: Traditional proportional-integral-derivative control
• LQR Control: Linear Quadratic Regulator for optimal control
• MPC Control: Model Predictive Control for predictive optimization

System Architecture

Key Components

1. Sensor Fusion (src/utils/sensor_fusion.py): Multi-sensor data integration
2. SLAM System (src/utils/slam_system.py): Real-time mapping and localization
3. BEV System (src/utils/bev_system.py): Top-down visualization
4. Control Algorithms (src/algorithms/control_algorithms.py): PID, LQR, MPC
5. Decision Algorithms (src/algorithms/decision_algorithms.py): Basic and advanced decision making
6. Avoidance Algorithms (src/algorithms/avoidance_algorithms.py): Lane change and pedestrian avoidance

Pedestrian Avoidance Features

• Multi-lane Detection: Detects pedestrians across all lanes
• Risk Assessment: Classifies collision risk levels (high/medium/low)
• Smart Stopping: Stops completely when all lanes are blocked by pedestrians
• Wait and Resume: Automatically resumes when path is clear

Installation & Deployment

Prerequisites

• Python 3.8+
• Pygame 2.6.1+
• NumPy

Quick Start

# Clone the repository
git clone <repository-url>
cd auto-simulation

# Install dependencies
pip install pygame numpy

# Run the simulation
python main.py

Controls

• 1-5: Switch edit modes (Road, Vehicle, Traffic Light, Obstacle, Pedestrian)
• Space: Pause/Resume simulation
• S: Toggle sensor visualization
• P: Toggle prediction visualization
• L: Switch traffic light to left-turn mode
• Right Click: Add waypoints for selected vehicle

Configuration

Modify src/core/constants.py to adjust simulation parameters:

• Screen dimensions
• Vehicle dynamics
• Sensor ranges
• Control parameters

Algorithm Customization

The system uses interface-based architecture for easy algorithm swapping:

Control Algorithms

# Switch between PID, LQR, or MPC
vehicle.control_algorithm = PIDControlAlgorithm()
vehicle.control_algorithm = LQRControlAlgorithm()  
vehicle.control_algorithm = MPCControlAlgorithm()

Avoidance Algorithms

# Choose avoidance strategy
vehicle.obstacle_avoidance_algorithm = LaneChangeAvoidanceAlgorithm()
vehicle.obstacle_avoidance_algorithm = PredictiveAvoidanceAlgorithm()
vehicle.obstacle_avoidance_algorithm = PedestrianAvoidanceAlgorithm()  # Default

Performance Features

• Real-time sensor fusion and object tracking
• Advanced pedestrian trajectory prediction
• Multi-lane collision risk assessment
• Adaptive control based on road conditions
• Extensible algorithm architecture

Development

The codebase follows modular architecture with clear separation of concerns:

• Entities: Vehicle, Road, TrafficLight, Obstacle, Pedestrian classes
• Algorithms: Control, decision, and avoidance algorithms
• Utils: Sensor fusion, SLAM, BEV visualization
• Core: Constants, enums, and base types

License

This project is for educational and research purposes.