Control Synthesis on TL Based Safe Reinforcement Learning
Control Synthesis on Temporal Logic-Based Safe Reinforcement Learning
Develop a novel methodology for multiple autonomous system task collaboration
Project Objectives
within two years
FORMAL MODEL
Developing a formal model using temporal logic in the collaborative task control policy – to provide reward for the RL agent, perform goal selection for the control Lyapunov function for the safe exploration. The formal model will be validated in a hybrid model checker tool.
SYSTEM CONSTRAINTS
Developing a finite automata augmented MDP framework to handle constraints and violation of the specification.
RL SYNTHESIS
Applying the above formal control framework and deep learning algorithm to a case study of multiple drone collaboration searching task in a blocked controlled environment.
OFFICE HOURS
Come Visit Me at
Mon - Fri: 9am - 11am
Sat: 10am - 2pm
Sun: Closed
any other time, email me at yujian.fu@aamu.edu