Data-driven situation awareness and multi-objective coordinated dispatch for renewable energy and storage

• Anomaly detection based on unsupervised learning ensures high-quality power data. • Hierarchical repair method enhances physical consistency of time-series power data. • K-means scenario extraction supports data-driven S-G-L-S situation awareness. • IMROA achieves multi-objective coordinated dispatch of renewable energy and storage. • AI-based integrated framework supports energy dispatch and sustainable power system. The rapid proliferation of renewable distributed energy (RDE) and energy storage device (ESD) introduces high-dimensional uncertainty and nonstationarity into modern hybrid energy systems, posing significant challenges for coordinated source-grid-load-storage (S-G-L-S) dispatch. To address this issue, a hybrid integrated dispatch framework is proposed that couples unsupervised learning with bio-inspired multi-objective optimization. On the perception side, a data-driven situation awareness module integrates physics-guided anomaly detection, scenario-based imputation, and K-means clustering to construct representative dispatch scenarios from real-world renewable and load datasets. On the decision-making side, an enhanced multi-objective remora optimization algorithm, which features adaptive parameter learning, elite solution memory, and multi-level evaluation, is developed to improve convergence and robustness under dynamic renewable environments. Case studies on IEEE benchmark systems demonstrate that the proposed framework significantly enhances dispatch quality, resource allocation efficiency, and renewable energy utilization compared with state-of-the-art methods. By integrating data-driven scenario modeling with adaptive optimization, the study underscores the role of artificial intelligence in advancing energy dispatch and supporting sustainable, low-carbon energy systems under uncertainty