Grid-Interactive Multi-Zone Building Control Using Reinforcement Learning with Global-Local Policy Search.
Xiangyu ZhangRohit ChintalaAndrey BernsteinPeter A. GrafXin JinPublished in: CoRR (2020)
Keyphrases
- policy search
- reinforcement learning
- continuous state
- control problems
- reinforcement learning algorithms
- optimal control
- dynamic programming
- continuous action
- reward function
- control policy
- control policies
- control strategies
- policy gradient
- function approximation
- model free
- state space
- partially observable markov decision processes
- function approximators
- markov decision problems
- control system
- learning algorithm
- markov decision processes
- multi agent
- np hard
- transfer learning
- real valued