RESEARCH OF SYSTEM WIDE INFORMATION MANAGEMENT SURVIVABILITY EVALUATION MODEL BASED ON BACK PROPAGATION NEURAL NETWORK

摘要

At present, the SWIM (System Wide Information Management) survivability has been a new research direction of the civil aviation. In this paper, constructs a PDRRDR survivability mode that is mainly in the dynamic services drift technology, and using the BP (Back Propagation) neural network to establish the evaluation model, with assessment and validation of the SWIM survivability. By using MATLAB software and through the reasonable sample data, realise the training and testing of neural network model, output error within the acceptable range. The evaluation model has a strong adaptability and fault-tolerant ability, therefore provides a highly efficient, accurate scientific means and tools for the complex SWIM survivability evaluation.