To establish a standardized framework for simulation systems and associated data system research, a diversified time-series dataset comprising both conventional and anomalous data subsets was constructed. Leveraging multi-source data including public datasets, mechanism-based simulation results, project-desensitized data, and large-scale model-generated data, the study employed a hybrid AI-driven methodology integrating mechanistic, data-driven, and knowledge-based approaches. This enabled demand-oriented generation of diversified data, thereby facilitating the integration of data resources essential for information-energy system research.