Abstract
For Unmanned Aircraft Systems, more advanced autonomous capabilities are being pushed to use in the near future, in order to relieve human fatigue and workload. Autonomous behavior requires intelligent agents to adapt their behavior to unexpected situations and examine their past experiences to increase their future mission
performance. To achieve adaptive behavior and decision making capabilities this study investigated the application of different Reinforcement Learning Models (Monte Carlo (MC) and Temporal Difference (TD)) and the effects of sensor sensitivity, as modeled through Signal Detection Theory (SDT), on the ability of RL algorithms to accomplish a target localization task. Three levels of sensor sensitivity are simulated and compared to the results of the same system using a perfect sensor. An evaluation of the system is performed using multiple metrics, including episodic reward curves and the time taken to locate all targets. Statistical analyses are employed to detect significant differences in the comparison of steady-state behavior of different systems. The study found that in general, TD models perform better than MC models, and sensor sensitivity has a significant effect on the agent performance as well.
Original language | American English |
---|---|
Title of host publication | Proceedings of the 2015 Industrial and Systems Engineering Research Conference |
State | Published - Jan 2015 |
Keywords
- unmanned aircraft systems
- human fatigue
- workload relief
- reinforcement learning models
- sensor sensitivity
Disciplines
- Graphics and Human Computer Interfaces
- Experimental Analysis of Behavior