Robust coordinated resilience enhancement strategy for communication networks of power and thermal cyber-physical systems considering decision-dependent uncertainty

The extensive deployment and application of communication equipment have enabled power and thermal cyberphysical systems (PTCPSs) to have highly flexible response measures while increasing the threat to communication networks from natural disasters. Due to the tight coupling of communication networks and energy networks, this threat will inevitably introduce risks and challenges for the operation of PTCPSs. To address these risks and challenges, this paper presents a robust coordinated resilience enhancement strategy model for PTCPSs that considers decision-dependent uncertainty (DDU) and mixed-integer recourse (MIR). To further reduce the investment and operation costs, path coordination and expansion strategies for power and thermal communication networks are considered based on the traditional hardening strategy. Then, to solve the nonlinear problems caused by path constraints and reduce the model dimensionality, a communication flow constraint parameterization method based on Yen's algorithm is proposed to convert the path constraints into parameter matrices. Due to the existence of DDU sets and MIR in the model, traditional decomposition methods are inapplicable. To solve this problem, a relaxed column and constraint generation and robust cut (RCCG-RC) method is presented with a detailed derivation, which adopts the ideas of linear relaxation and fixed-variable methods. Besides, strong duality theory and Karush-Kuhn-Tucker conditions are utilized, and several subproblems are reformulated as linear subproblems. This method converges well after a limited number of iterations. The case studies demonstrate the rationality and effectiveness of the proposed model and method, which can effectively improve the economics of enhancing PTCPS resilience. Compared with the benchmark of the traditional hardening strategy, the proposed method reduces the investment cost by up to 52.78%.