This is the project report of 2021 fall course CUHK IERG5350.
The report paper can be found here.
The interaction of our agent with the power environment is showed as:
The learning curve of TD3 algorithm and the cost function during training:
Compared with traditional control methods, our RL control scheme achieves the voltages profiles as:
