kundtx / lfd2022-comments

0 stars 0 forks source link

Learning from Data (Fall 2022) #27

Open kundtx opened 1 year ago

kundtx commented 1 year ago

http://8.129.175.102/lfd2022fall-poster-session/4.html

HaoJun-Zhang99 commented 1 year ago

G12 Haojun Zhang: Impressive work! Wanna to know what strategy does the Multi-agent learn to manage the energy system?

ljh430 commented 1 year ago

G38 Jiahao Lai:What's the shared information between different agents?From the figure it seems that variable c and e is shared,what does that mean?

NYZTTTT commented 1 year ago

@HaoJun-Zhang99 G12 Haojun Zhang: Impressive work! Wanna to know what strategy does the Multi-agent learn to manage the energy system?

G4 Zihao Ni: Agents learn to control the storage batteries charging/discharging in buildings to minimize the electricity demand from the grid. There's a dim-1 continuous action space for each agent, of which positive value means to charge the battery with that exact power(in kWh).

NYZTTTT commented 1 year ago

@ljh430 G38 Jiahao Lai:What's the shared information between different agents?From the figure it seems that variable c and e is shared,what does that mean?

G4 Zihao Ni: c^i is the capacity dispatched to each building i , which means the credibility of the predicition made by the ith agent. e^i means the predicted electricity consumption made by the ith agent. The shared information is sigma(c^i) and sigma(e^i), which means the accumulated credibility and predicted value of accumulated electricity consumption.