A Novel Optimal Energy Management and Operation in Sports Facilities

This paper introduces a Deep Learning–Enhanced Energy Management Framework for optimizing solar integration and dynamic load scheduling in sports facilities. At its core, a Deep Temporal Convolutional Network (D-TCN) performs multi-step forecasting of solar generation and energy demand, leveraging causal and dilated convolutions to efficiently capture long-term temporal patterns while maintaining low computational latency. Integrated with a Smart Energy Management System (SEMS) for real-time HVAC, lighting, and battery control, the framework employs a Modified Aquila Optimization Algorithm (M-AOA) enhanced with adaptive penalties and fuzzy logic for dynamic battery dispatch. This hybrid approach effectively minimizes grid dependence and operational costs while preserving occupant comfort. Simulation results reveal a 24.7% reduction in total energy consumption and a 41.2% increase in solar utilization. The scalable IoT-based architecture ensures seamless integration with existing smart infrastructure, supporting national carbon neutrality and demand-side management objectives for sustainable urban energy systems.