In the context of high renewable energy penetration, the demand for power grid flexibility regulation is growing, yet existing load-side flexibility resource research ignores spatio-temporal coupling effects, leading to fragmented mining, insufficient quantification adaptability and low aggregation efficiency. This paper proposes an adaptive mining and quantification method for load-side flexibility resources considering spatio-temporal coupling. First, an adaptive mining framework integrating multi-energy time patterns, regional spatial clustering and spatio-temporal coupling matching is constructed to locate resources accurately. In this framework, the ST-MLCM (spatio-temporal multi-layer chain model) and ST-GMM (spatio-temporal Gaussian mixture model) are mathematically extended with spatial coupling terms, enabling genuine joint modeling of temporal dynamics and spatial dependencies. Second, a quantification model combining AHP-entropy weight method and fuzzy multi-attribute decision-making with spatio-temporal dynamic correction is established to balance weights and heterogeneity. Third, a ST-GNN-based distributed aggregation method is designed to realize high-efficiency aggregation of large-scale resources. Experimental results verify the effectiveness of the method in resource mining, quantitative evaluation and aggregation application, providing technical support for spatio-temporal coordinated dispatching of new power systems.
Zheng et al. (Mon,) studied this question.