In order to improve the rationality and effectiveness of critical infrastructure security resource allocation, and solve the shortcomings of traditional methods in the balance of security resource allocation and protection effectiveness, an optimization model of critical infrastructure security resource allocation was constructed based on PT, ASD and Stackelberg game theory. Firstly, the ASD-Stackelberg game model was constructed to solve the optimal spatial allocation of security resources within critical infrastructure. Secondly, the PT theory was introduced to describe the bounded rationality characteristics of both attack and defense sides, and the optimal security resource investment threshold was determined. Finally, the nuclear power plant security system was taken as a case study. The verification and analysis results show that compared with the traditional empirical configuration method, the optimal resource allocation scheme output by this model has significantly improved the overall security effectiveness of nuclear power plants by 25.9%, and can accurately identify the optimal amount of security resource investment. This result not only confirms the effectiveness and feasibility of the model in solving the optimal configuration scheme, but also quantitatively establishes its significant superiority in enhancing security protection capabilities.