Storage Reliability of Military System Based on Periodic Test Data

摘要

This paper proposes a detailed framework to analyze and assess storage reliability of military system based on periodic test data and a theoretical and scientific data warehouse to realize the processes of storage reliability analysis. Traditional methods used for reliability prediction are not suitable for the periodic test dataset. Thus a new approach of storage reliability prediction based on least squares support vector machine and particle swarm optimization is introduced to resolve it. Numerical example shows that the new approach is much more rational and exactitude than polynomial regress.