Hou Shengren

icon company @TUDELFT icon location Delft, Netherlands icon location Email: [email protected] PhD candidate in EWI, TU Delft. Research Area: Reinforcement Learning based Energy Management, Power Market,

Results 3 comments of Hou Shengren

Parameters

hey, sry for the late reply. I just back from holiday. current_capacity is SOC, you are right that battery capacity should be kwh. At that moment, when I create this...

Some question about the MIP_DQN.py

This class is called after training is finished when doing real-time implementation. After training, you get a well trained Q network (critic network) and actor network. We don't use this...

State space issue

This doesn't impact the performance according to my experience, state normalization and denormalization is to stable the training by scaling the state into a magnitude that agents can easily handle....