Perception-based Co-evolutionary Reinforcement Learning for UAV Sensor Allocation

摘要

In this project, we have formulated the problem of sensor allocation in a team of UAVs within a mathematical programming framework. A Perception- based reasoning approach based on co-evolutionary reinforcement learning was developed for jointly addressing sensor allocation on each individual UAV and allocation of a team of UAVs in the geographical search space. An elaborate problem setup was simulated and experimented with, for testing and analysis of this framework using the Player-Stage multi-agent simulator. This simulator was developed jointly at the USC Robotics Research Lab and HRL Labs.The experimental results demonstrated a very strong performance of our methodology for UAV sensor allocation problem domains. Our results indicate that not only it is feasible to use perception-based reinforcement learning for this problem but it is an adequate solution for many typical UAV teams.