This paper proposes a game-theoretic optimisation framework for designing attack and defence strategies in hybrid AC/DC cyber-physical power systems (CPPS). The proposed framework employs a three-layer planning hierarchy. The upper layer introduces a novel target importance index combining complex network theory and power flow entropy. The middle layer incorporates attack success probability and defines a multidimensional damage index based on topology reduction

load loss and power flow deviation. The lower layer develops an optimal dispatch model integrating economic cost and security penalties

addressing both normal operations and emergency responses. Simulations on a modified IEEE 39-bus system demonstrate that the framework improves defence efficiency and system resilience

whilst revealing a diminishing marginal effect in defence investment. The analysis further quantifies how defence sensitivity parameters affect attack suppression

supporting resource allocation decisions. This work offers a scalable solution for enhancing CPPS security against cyber-physical threats

with potential extensions to adaptive and cooperative defence strategies.