Design And Implementation Of BP Neural Network Based Risk Management Systems For Hybrid Unit

. Currently, the training speed of flight personnel can hardly catch the rapid development of civil aviation enterprise, in order to solve the problems of the shortage of flight personnel and introduce into advanced management experience, a large number of foreign pilots have been recruited by domastic airline companys. With the join of these foreign flight personnel, the flight crew become mixed of persons from different country, a security risk arises due to cultural differences. So, it is urgent to build a risk management information system special for mixed flight crew by using advanced technology. It is able to provide a risk analysis and management platform for risk managers so as to reduce the potential risk of cultural differences, at the same time,it helps to reduce the cost of management and realize the effective manegement of the mixed flight crew.This project aimed at addressing the complex problem of risk management for mixed flight crew. It makes full use of the advanced computer technologies such as BP neural network technology, knowledge base, Mat Lab and so on to build a BP neural network based risk management system for mixed flight crew, which can mine and analyze the risk information for various mixed flight crew and propose scientific means of effective management in order to reduce the risk posed by cultural differences and the management costs.The main contents of this paper include the following aspects: Firstly, it analyzes domestic and international situation of risk information management for mixed flight crew and the relationship between the factors of cultural differences and risk, as well as the status and trend of the BP neural network technology,based on it to prove the feasibility of this project, choose the suitable new technology and determine the project demand; Secondly, it provides a detailed study to the risk caused by cultural differences, establishs the relationship between risk factors and cultural differences, constructs an appropriate model by using the BP neural network technology and conducts the overall design of specific projects, finishes functional module division and detailed design; Thirdly, it tests the functionality and performace of the developed BP neural network based risk information management system for mixed flight crew, summarizing the development and implementation of the system and discussing the future work.By conducting a series of functional tests to the BP neural network based risk information management system for mixed flight crew, it was verified in this work that the designed and implemented system meets the basic design requirements, it realizes the scientific and efficient management of the risk information for mixed flight crew. From the practical application point of view, it basically satisfies the requirements of the desired design and is worth for application to some extent.