Efficiency improvement in energy consumption: A novel deep learning based model for leading a greener Economic recovery

Abstract

This research focuses on elevating energy consumption efficiency through an innovative predictive model designed for urban industrial and residential sectors. Leveraging the Deep Q Network (DQN) framework, the study introduces the Dueling DQN architecture, a breakthrough approach that separates state value and advantage function estimations. This segregation significantly enhances learning efficiency, providing a more precise representation of the state's value and, consequently, optimizing energy consumption predictions. In tandem with architectural advancements, the research proposes a unique fuzzy Particle Swarm Optimization (PSO) approach to fine-tune DQN setting parameters during training. This integration of fuzzy logic into PSO ensures adaptability and robustness, addressing challenges associated with diverse environmental and operational conditions. For the goal of efficiency improvement in energy consumption, this study not only introduces a state-of-the-art solution for energy consumption prediction and optimization but also aligns with the broader goal of fostering a greener economic recovery. The model's promising results signify a significant step toward more sustainable and resource-efficient urban development, contributing to a resilient and environmentally conscious economic future. This research lays the foundation for a transformative approach to energy consumption, vital for steering economic recovery toward a greener and more sustainable trajectory. The simulation results advocate the high performance of the proposed deep learning prediction model.