| Abstract |
The growing importance of renewable energy sources in achieving net -zero emissions and landscape design has made accurate short-term load forecasting (STLF) a critical component of the power sector. STLF plays a pivotal role in smart grid technologies, impacting power production scheduling, load management, procurement strategies, facility maintenance, and contractual evaluations. This study focuses on the development of a precise and rapidly convergent STLF model within the context of digital twin technology for renewable energy integration. To enhance prediction accuracy, we employ two widely recognized methods: mutual data -enabled attribute selection and a modified flower pollination algorithm (MFPA) aimed at error reduction. Additionally, we adapt heuristic algorithms and training procedures for artificial neural networks to expedite the convergence rate of our comprehensive prediction approach. Through extensive simulations, our findings demonstrate that this proposed prediction method achieves an impressive accuracy rate of 99.5 percent while significantly reducing the average runtime, by 52.38 percent, compared to traditional bilevel forecasting approaches. This research showcases the potential of digital twin modeling to optimize the integration of renewable energies into smart city infrastructures, advancing the path towards a sustainable, net -zero emissions future. |
