YangyangFu/autonomous-driving-book

A book for autonomous driving

Autonoumous Driving: Beginner to Master

This book aims to help beginners to learn autonomous driving from coding excercises. The content is organized in a bottom-up approach, by introducing local longitudinal and lateral PID controller first, and all the way up to perception. All topics are demosntrated through closed-loop simulation using CARLA. I found this approach is easier for students in mechanical/electrical engineering to start with.

The covered topics are:

• 1 Introduction
• 2 Vehicle Dynamics
  • 2.1 Kinematic Bicycle Model
  • 2.2 Dynamic Bicycle Model
• 3 Trajectory Tracking
  • 3.1 Longitudinal Control
    • 3.1.1 PID Controller
  • 3.2 Lateral Control
    • 3.2.1 Basic Concepts
    • 3.2.2 Pure Pursuit Controller
    • 3.2.3 Stanley Controller
    • 3.2.4 LQR Controller
    • 3.2.5 MPC Controller
  • Combined Control
    • MPC Controller
• Motion Planning
  • Route Planning
  • Behavior Planning
  • Motion Planning
• Motion Prediction
  • Sensor Fusion
  • Multi-target Tracking
  • Motion Prediction
• Simultenanous Localization and Mapping (SLAM)
  • Localization
  • Mapping
  • SLAM
• Perception
  • Camera 3D Object Detection
  • Lidar 3D Object Detection
  • Multi-sensor Object Detection

How to Start

1. Run Carla server in docker. Make sure you have docker installed. Direct to the root folder of this repo and run the following command in a terminal:

   For Ubuntu,

       bash ./run_server/run_carla_0.9.14.sh

   For Windows,

       ./run_server/run_carla_0.9.14.bat

2. Run visualization. Open a new terminal and type:

   bash run_visualization/run_carla_viz.sh

   Then open a browser and go to localhost:8000.

3. Open a new terminal to run Carla client:

   For Ubuntu,

   source setup_pythonpath.sh 
   cd ./book/3-trajectory-following
   python test_lateral_control.py

Contribution

If you are interested in any of the above topics, please feel free to contact me at fuyy2008@gmail.com