Real-time acquisition of system operating states with high temporal and spatial resolution is fundamental to achieving online dynamic security assessment (DSA) for modern power systems. However

traditional state estimation (SE) research faces challenges in data availability

such as insufficient spatiotemporal resolution of heterogeneous measurements. Additionally

methodological limitations persist

including the constraints of single-mechanism models

the difficulty of balancing timeliness and reliability in data-driven approaches

and the challenge of analytically constructing system dynamic models. To address these issues

this paper first reviews the previous research in state estimation and summarizes the key challenges in achieving high-resolution estimation. Then

based on the principle of “physics-based and data-driven combination”

a high spatial-temporal resolution SE framework is designed for dynamic security analysis in modern power systems. The paper proposes techniques for constructing

training

improving

and applying high-resolution SE models

considering the specific characteristics of different classical hybrid-driven models. Finally

case studies validate the effectiveness of the proposed high spatiotemporal resolution state estimation techniques.