A reactive power optimization partially observable Markov decision process with data uncertainty using multi-agent actor-attention-critic algorithm

•A novel POMDP modelling method for reactive power optimization of the ADN which fills the gap in the solution to the optimization problem with significant data uncertainty.•Introduce a technique for transforming the POMDP model into the belief state which can be easily solved by the RL algorithm.•U...

Celý popis

Podrobná bibliografie

Vydáno v:	International journal of electrical power & energy systems Ročník 147; s. 108848
Hlavní autoři:	Gu, Yaru, Huang, Xueliang
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	Elsevier Ltd 01.05.2023
Témata:	Data uncertainty Multi-agent reinforcement learning algorithm Partially Observable Markov Decision Process(POMDP) Reactive power dispatching optimization model Voltage control Reactive power dispatching optimization model Multi-agent reinforcement learning algorithm Partially Observable Markov Decision Process(POMDP) Voltage control Data uncertainty
ISSN:	0142-0615, 1879-3517
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!