Abstract
Power systems are encountering unprecedented stability, resilience, and intelligent operation challenges against the backdrop of high-penetration renewable energy integration and deep energy system transformation. As a core pillar for building next-generation power systems, smart perception and collaborative control systems rapidly evolve from traditional centralized and static dispatching modes toward architectures characterized by real-time sensing, distributed decision-making, and cognitive control. This paper reviews the developmental trajectory and overarching technological framework of such systems, focusing on key technological pathways including multi-source heterogeneous state perception, edge-intelligent processing, AI-driven autonomous decision-making mechanisms, and multi-agent collaborative control. In light of persistent challenges such as insufficient data reliability, limited generalization of control strategies, and lagging resilience assurance mechanisms, this paper proposes that future efforts should prioritize the development of a cognitively enabled smart grid core system with reasoning, self-healing, and cross-scale coordination capabilities. The paper aims to promote the deep integration of digitalization, intelligence, and resilience in power grid evolution.