Journal-2: A Dynamic pricing demand response algorithm for smart grid: Reinforcement learning approach

Journal-2: A Dynamic pricing demand response algorithm for smart grid: Reinforcement learning approach

Journal: Applied Energy

Impact factor: 7.900

中科院分区:一区

DOI: https://doi.org/10.1016/j.apenergy.2018.03.072

Although the paper is accepted, there are two issues to consider when you apply the reinforcement learning in decision making problems.

1. We should carefully define the reward function. The quality of the reward function will directly affect the learning strategy.

2. The convergence time of reinforcement learning based algorithms is highly dependent on action number, which means that arbitrary increase in actions could lead to longer convergence time and even wrong results.

PDF version:

1-s2.0-S0306261918304112-main

发表评论

电子邮件地址不会被公开。 必填项已用*标注