Abstract
Introduction: In the context of the evolving energy landscape, the efficient integration of energy storage systems (ESS) has become essential for optimizing power system operation and accommodating renewable energy sources.Methods: This study introduces LoadNet, an innovative approach that combines the fusion of Temporal Convolutional Network (TCN) and Gated Recurrent Unit (GRU) models, along with a self-attention mechanism, to address the challenges associated with ESS integration in power system operation. LoadNet aims to enhance the management and utilization of ESS by effectively capturing the complex temporal dependencies present in time-series data. The fusion architecture of TCN-GRU in LoadNet enables the modeling of both short-term and long-term dependencies, allowing for accurate representation of dynamic power system behaviors. Additionally, the incorporation of a self-attention mechanism enables LoadNet to focus on relevant information, facilitating informed decision-making for optimal ESS operation. To assess the efficacy of LoadNet, comprehensive experiments were conducted using real-world power system datasets.Results and Discussion: The results demonstrate that LoadNet significantly improves the efficiency and reliability of power system operation with ESS. By effectively managing the integration of ESS, LoadNet enhances grid stability and reliability, while promoting the seamless integration of renewable energy sources. This contributes to the development of a more sustainable and resilient power system. The proposed LoadNet model represents a significant advancement in power system management. Its ability to optimize power system operation by integrating ESS using the TCN-GRU fusion and self-attention mechanism holds great promise for future power system planning and operation. Ultimately, LoadNet can pave the way for a more sustainable and efficient power grid, supporting the transition to a clean and renewable energy future.