Sustainability, Journal Year: 2025, Volume and Issue: 17(8), P. 3584 - 3584

Published: April 16, 2025

Reliable and precise joint probabilistic forecasting of wind solar power is crucial for optimizing renewable energy utilization maintaining the safety stability modern systems. This paper presents an innovative model designed to address spatiotemporal output challenges. Leveraging a multi-network deep learning framework, integrated temporal convolutional network feature extraction, neural spatial analysis, attention mechanism focus enhancement, thereby capturing complementarity power. It also incorporated quantile regression-based uncertainty quantification technique, contributing reliable predictions. A farm two farms in China were used as case study. Comparison results between proposed ten established models demonstrated its superior performance both deterministic predictions, offering valuable insights sustainable resilient system operation.

Language: Английский