• A Swathi
• Noothi Sravan Kumar
• S V N SAI RAVI KIRAN
• SANAGAPALLI VENKATA NAGA SAI SRIYA

With the fast-evolving smart grids and energy-intensive infrastructure energy demand and forecasting and early detection of irregular patterns for efficient energy management. This paper proposes DAIEN (Dual-Stream Adaptive Edge Intelligence Network), A deep learning architecture designed for smart energy forecasting and demand irregularity detection. The model uses a dual stream approach combining time-related energy usage patterns with features using an attention enhanced Gated Recurrent Unit (GRU) network optimized for edge computing conditions. The proposed system performs forecasting and simultaneously identifying abnormal patterns such as sudden spikes and energy failures. Experimental Evaluations are conducted on the smart meter datasets which demonstrates improved prediction accuracy and anomaly detection performance compared to traditional forecasting models.

Smart Energy Forecasting, Demand Irregularity Detection, Edge Computing, Gated Recurrent Unit (GRU), Attention Mechanism, Smart Grid Analytics, Deep Learning.

"DAEIN: Dual-Stream Attention-GRU for Energy Forecasting and Demand Irregularity Detection ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 9, page no.d107-d112, September-2026, Available :http://www.jetir.org/papers/JETIR2609313.pdf

"DAEIN: Dual-Stream Attention-GRU for Energy Forecasting and Demand Irregularity Detection ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 9, page no. ppd107-d112, September-2026, Available at : http://www.jetir.org/papers/JETIR2609313.pdf