diff --git a/AgentPPO/actor.pth b/AgentPPO/actor.pth index 5109f84..857688d 100644 Binary files a/AgentPPO/actor.pth and b/AgentPPO/actor.pth differ diff --git a/AgentPPO/loss_data.pkl b/AgentPPO/loss_data.pkl index d28969a..a019fe1 100644 Binary files a/AgentPPO/loss_data.pkl and b/AgentPPO/loss_data.pkl differ diff --git a/AgentPPO/reward_data.pkl b/AgentPPO/reward_data.pkl index 1ba7e24..d420577 100644 Binary files a/AgentPPO/reward_data.pkl and b/AgentPPO/reward_data.pkl differ diff --git a/DDPG.py b/DDPG.py index 9612469..5057afe 100644 --- a/DDPG.py +++ b/DDPG.py @@ -105,7 +105,7 @@ if __name__ == '__main__': record = test_one_episode(env, agent.act, agent.device) eval_data = pd.DataFrame(record['system_info']) eval_data.columns = ['time_step', 'price', 'netload', 'action', 'real_action', 'soc', 'battery', 'gen1', 'gen2', - 'gen3', 'unbalance', 'operation_cost'] + 'gen3', 'temperature', 'irradiance', 'unbalance', 'operation_cost'] if args.save_test_data: test_data_save_path = f'{args.cwd}/test_data.pkl' with open(test_data_save_path, 'wb') as tf: diff --git a/PPO.py b/PPO.py index 21c4171..35d8484 100644 --- a/PPO.py +++ b/PPO.py @@ -130,7 +130,7 @@ class AgentPPO: for i in range(target_step): action, noise = self.select_action(state) state, next_state, reward, done, = env.step( - np.tanh(action)) # here the step of cut action is finally organized into the environment. + np.tanh(action)) # the step of cut action is finally organized into the environment. trajectory_temp.append((state, reward, done, action, noise)) if done: state = env.reset() @@ -371,7 +371,7 @@ if __name__ == '__main__': record = test_one_episode(env, agent.act, agent.device) eval_data = pd.DataFrame(record['system_info']) eval_data.columns = ['time_step', 'price', 'netload', 'action', 'real_action', 'soc', 'battery', 'gen1', 'gen2', - 'gen3', 'unbalance', 'operation_cost'] + 'gen3', 'temperature', 'irradiance', 'unbalance', 'operation_cost'] if args.save_test_data: test_data_save_path = f'{args.cwd}/test_data.pkl' with open(test_data_save_path, 'wb') as tf: diff --git a/SAC.py b/SAC.py index 77852ff..07eb753 100644 --- a/SAC.py +++ b/SAC.py @@ -109,7 +109,7 @@ if __name__ == '__main__': record = test_one_episode(env, agent.act, agent.device) eval_data = pd.DataFrame(record['system_info']) eval_data.columns = ['time_step', 'price', 'netload', 'action', 'real_action', 'soc', 'battery', 'gen1', 'gen2', - 'gen3', 'unbalance', 'operation_cost'] + 'gen3', 'temperature', 'irradiance', 'unbalance', 'operation_cost'] if args.save_test_data: test_data_save_path = f'{args.cwd}/test_data.pkl' with open(test_data_save_path, 'wb') as tf: diff --git a/TD3.py b/TD3.py index 5dd2e6f..050a2af 100644 --- a/TD3.py +++ b/TD3.py @@ -105,7 +105,7 @@ if __name__ == '__main__': record = test_one_episode(env, agent.act, agent.device) eval_data = pd.DataFrame(record['system_info']) eval_data.columns = ['time_step', 'price', 'netload', 'action', 'real_action', 'soc', 'battery', 'gen1', 'gen2', - 'gen3', 'unbalance', 'operation_cost'] + 'gen3', 'temperature', 'irradiance', 'unbalance', 'operation_cost'] if args.save_test_data: test_data_save_path = f'{args.cwd}/test_data.pkl' with open(test_data_save_path, 'wb') as tf: diff --git a/environment.py b/environment.py index 078d9b0..2d8a001 100644 --- a/environment.py +++ b/environment.py @@ -175,7 +175,7 @@ class ESSEnv(gym.Env): # process_elements(pv, lambda x: x, self.data_manager.add_pv_element) process_elements(price, lambda x: max(x / 10, 0.5), self.data_manager.add_price_element) - process_elements(load, lambda x: x * 5, self.data_manager.add_load_element) + process_elements(load, lambda x: x * 3, self.data_manager.add_load_element) process_elements(irradiance, lambda x: x, self.data_manager.add_irradiance_element) process_elements(temperature, lambda x: x - 273.15, self.data_manager.add_temperature_element) process_elements(wind, lambda x: x, self.data_manager.add_wind_element)