Analysis of Risk-Based Airport Passenger Classification with PSO-BP Neural Network
Zhenwu Zhao, Chenchen Zhang, Dongdan Guo · 2020
In order to reduce passengers' waiting time, improve the quality of security inspection, minimize the cost of service resource and maximize the utilization of security resource, the index system of risk-based airport passenger classification was constructed to identify passenger risk levels. Firstly, the index system of risk-based airport passenger classification was initially constructed through literature review and expert consultation. Secondly, the index system of risk-based airport passenger classification is optimized with the help of relevant professionals' questionnaires and SPSS22.0 software. Finally, BP neural network classifier is selected and which was optimized by PSO algorithm. The classifier was trained in simulations and tested for classification effects by using the sample data of questionnaire survey. Test results show that the index system of risk-based airport passenger classification can be obtained to measure passenger risk degree from five aspects, natural condition, occupation status, economic condition, credit situation and flight condition. The effectiveness of the index system and the BP neural network classifier were verified by the classification results of 5 passengers through the classifier.