Evaluation study on comprehensive efficiency of physical protection system based on neural network

[Background] At present, most of the physical protection systems of nuclear facilities at home and abroad use the path analysis method to evaluate the effectiveness. The performance of the system is evaluated by calculating the cut off probability, hence the evaluation index is single and cannot reflect factors such as the system reliability, the command strategy, the information security design and other factors. [Purpose] This study aims to analyze the comprehensive effectiveness of the physical protection system by using neural network method. [Methods] The comprehensive effectiveness evaluation index system of the physical protection system was established by using the analytic hierarchy process (AHP), and the comprehensive effectiveness evaluation model of the physical protection system for nuclear facilities was established by using the composite ability of the nonlinearity of the BP (back propagation) neural network. The comprehensive effectiveness of the physical protection system for the marine nuclear power platform was evaluated. [Results and Conclusions] The evaluation results show that the BP neural network evaluation model converges fast, avoids the shortcomings of the majority of sample concentration effects in the AHP evaluation process, and has better adaptability and stability. It provides a kind of field for physical protection effectiveness evaluation. (authors)