Multi-stage resilience enhancement method for integrated power-thermal energy systems considering coordinated cyber-physical attacks

With the increasing trend toward intelligence and multi-energy coordination in integrated energy systems, the power, thermal, and communication subsystems are becoming highly interconnected, forming a typical Power and Thermal Cyber-Physical System (PTCPS). While such integration improves operational efficiency, the complex coupling structure significantly increases the system’s vulnerability under cyber-physical coordinated attacks. Existing studies have primarily focused on resilience enhancement at the physical layer, lacking systematic modeling of cyber-physical interactions. Moreover, most strategies are limited to either the pre-disturbance or post-disturbance stage, failing to achieve effective resource coordination and rapid system recovery. To address these challenges, this paper proposes a multi-stage resilience enhancement approach for PTCPS, which comprehensively considers both physical coupling between power and thermal systems and interaction mechanisms at the cyber layer. On this basis, a full-cycle resilience enhancement framework is developed, covering both pre-disturbance defense and post-disturbance recovery. In the pre-disturbance stage, defense resources are optimally allocated across cyber and physical layers to improve resistance to cross-layer coordinated attacks. In the post-disturbance stage, existing multi-dimensional emergency resources are fully utilized—including emergency wireless communication supported by thermal base stations, short-term decoupled operation of electric and thermal subsystems, and cross-layer coordinated repair scheduling—to enable efficient resource coordination and accelerated restoration. A case study is conducted on a PTCPS built from a modified IEEE 39-bus power system and a 32-node thermal system. Comparative simulation results demonstrate that the proposed method outperforms benchmark strategies under cyber-physical coordinated attack scenarios, significantly enhancing both load preservation and recovery efficiency, and thereby improving overall system resilience.