-
Notifications
You must be signed in to change notification settings - Fork 6
Data_Source_Layer
The Data Source Layer in the RL-ADN framework provides the primary data required to simulate the environment. This includes historical load data, renewable energy generation data, and market price data. These data sources are crucial for creating realistic and varied scenarios for training DRL agents.
The data related to active and reactive power demand (( p^D_{i,t} ), ( q^D_{i,t} )) are essential for simulating the load on the distribution network. These data points are typically collected from historical records and are used to model the consumption patterns of different nodes within the network over time.
Electricity price (( \rho_t )) data is crucial for modeling the economic aspects of energy storage and distribution. This data helps in understanding the cost implications of different actions taken by the DRL agent, such as charging or discharging the energy storage systems (ESSs).
Renewable power generation data (( p^R_{i,t} ), ( q^R_{i,t} )) captures the contribution of renewable energy sources such as solar and wind to the distribution network. This data is important for modeling the variability and intermittency associated with renewable energy sources.
The Data Manager plays a crucial role in handling the time-series data for specific epochs (( \mathcal{T}, t\in \mathcal{T} )). Previous research approaches to data management have been case-specific and labor-intensive, adding complexity and potential data quality issues. RL-ADN adopts a streamlined approach, standardizing various data preprocessing tasks, and ensuring data integrity and efficient handling. The workflow of the Data Manager is detailed in Appendix~\ref{sec_data_manager_workflow}.
In RL-ADN, the Data Augmentation module enhances the robustness and generalizability of the trained policy by artificially expanding the diversity of the historical time-series data. This module generates synthetic time-series data, capturing the stochastic nature of load in the power system and reflecting realistic operational conditions. The module interacts with the Data Manager to retrieve the necessary preprocessed data and then applies its augmentation algorithms to produce an augmented dataset. The output is a synthetic yet realistic dataset that reflects the variability and unpredictability inherent in distribution network systems. This enriched dataset is crucial for training RL agents, providing them with a diverse range of scenarios to learn from and ultimately resulting in a more adaptable and robust decision-making policy. The workflow of the Data Augmentation module is described in Appendix~\ref{sec_data_augmentation_workflow}.
The Data Source Layer is foundational to the RL-ADN framework, providing the essential data required to create realistic and varied training scenarios for DRL agents. By standardizing data preprocessing and incorporating advanced data augmentation techniques, the RL-ADN framework ensures high-quality, diverse datasets that enhance the training process and the resulting policies' effectiveness.
For more detailed information on the workflow and data management, refer to the respective appendices and sections in this Wiki.