Online Reinforcement Learning Control by Direct Heuristic Dynamic Programming: From Time-Driven to Event-Driven.
Qingtao ZhaoJennie SiJian SunPublished in: IEEE Trans. Neural Networks Learn. Syst. (2022)
Keyphrases
- event driven
- dynamic programming
- reinforcement learning
- optimal control
- control problems
- state space
- markov decision processes
- optimal policy
- real time
- information delivery
- robot control
- online learning
- control system
- dynamic programming algorithms
- lagrangian relaxation
- function approximation
- linear programming
- publish subscribe
- greedy algorithm
- infinite horizon
- control strategies
- control policy
- optimal solution
- locally optimal
- robotic control
- reinforcement learning algorithms
- markup language
- information services
- single machine
- feasible solution
- relational databases