Bridging Physics-Informed Neural Networks with Reinforcement Learning: Hamilton-Jacobi-Bellman Proximal Policy Optimization (HJBPPO).
Amartya MukherjeeJun LiuPublished in: CoRR (2023)
Keyphrases
- neural network
- reinforcement learning
- function approximators
- optimal policy
- control problems
- function approximation
- approximate dynamic programming
- action selection
- hamilton jacobi bellman
- optimal control
- model free
- learning algorithm
- fuzzy logic
- state space
- markov decision process
- policy iteration
- stochastic control
- rl algorithms
- asymptotically optimal
- temporal difference
- action space
- step size
- evolutionary algorithm
- markov decision processes
- policy gradient
- artificial neural networks
- actor critic
- fuzzy systems
- learning problems