Risk Assessment of Aviation Maintenance Error Based on Set Pair Analysis and BP Neural Network
Ying Duan, Lingling Yuan · Proceedings of the international Asia conference on industrial engineering and management innovation · 2015
According to the complex association of the aviation maintenance error, using the SHEL model to divided the content of maintenance error into four subsystems including human-hardware subsystem, human-software subsystem, human-environment subsystem, human-human subsystem, and establishing a hierarchical model of ladder structure based on this. Considering the uncertainty and complexity of each subsystem, first establishing a set of risk assessment models of maintenance error through the set pair analysis (SPA), and dealing with date from same-indefinite-contrary points in order to acquire the degree of connection, and combining with AHP analysis method to determine subsystem’s weight,. Then each component values of the connection’s degree as input layer of BP neural network, simulating the maintenance error comprehensive risk by using BP neural network method. Finally, taking an aviation maintenance enterprise for an example, the network model is trained and tested from the practical data, and the quantitative results indicate the effectiveness and feasibility of the proposed method, which can provide a reasonable decision support for aviation maintenance management.