Embedded Files
Group 18, EECS 206B, Spring 2020
Demonstrating ILQ games for safety-critical multi-Agent Autonomous systems
Demonstrating ILQ games for safety-critical multi-Agent Autonomous systems
Some Context:
This project is the result of an ongoing research collaboration with David Fridovich-Keil, an early adopter of the ILQ-games approach for tackling safety-critical autonomous systems. The work presented here builds on his work developing a full featured ILQ-game solver.
ILQ Games
ILQ Games
We discuss the benefits of Iterative Linear Quadratic games in online real-time solution of multi-agent games.
Traffic Scenario
Traffic Scenario
We pose a realistic and challenging traffic scenario that can be solved using ILQ-games.
Traffic Simulation
Traffic Simulation
We formulate the traffic scenario as an ILQ game, solve, and simulate the results over a fixed-horizon
Real-time, Online 3D Simulation
Real-time, Online 3D Simulation
We develop an infrastructure for the use of ILQ-games to solve and simulate complex problems online in real-time using ROS, C++, and RVIZ.
For more details, read the paper written for this project below!
For more details, read the paper written for this project below!
206B_final report.pdf
Questions?
Questions?
Contact mkestrada2@berkeley.edu to get more information on the project
Code available by request
Page updated
Google Sites
Report abuse