Predictive Fault Detection for Missile Defense Mission Equipment and Structures

Abstract

Equipment failures in defense systems result in loss or reduction of operational capability, impacting system readiness. Faults in critical equipment can impact system performance and reliability, as can structural failures. Predictive fault detection (PFD) provides prognostic capability to identify components and internal structures that exhibit either degradation or increased variability of parameters, in advance of actual faults occurrences. It has the power to reduce facility down times significantly, enhancing system availability and reducing maintenance costs. PFD enables maintainers to take corrective action before actual failures occur, thus improving system availability and reducing maintenance costs for the covered facility. The authors have applied an integrated approach to PFD, including multiple-phenomenology sensing and data acquisition; generalized multivariate statistical treatment of waveform shapes, continuous variable values, and discrete event occurrences; and maintenance decision-making techniques that balance performance and cost factors. An engineering prototype PFD product has been developed and tested under an MDA-sponsored Small Business Innovative Research project sponsored by the Missile Defense Agency.