Autonomous Satellite Threat Processing and Response

ABSTRACT: The proposal considers a model based approach to detecting anomalous events and/or emerging threats to a satellite. The basis of the approach is the development of non-parametric Bayesian techniques for learning a model of satellite health. This model can then be used as the basis for detecting when the satellite enters anomalous states corresponding, for instance, to a component failure, the impact of space weather, or a deliberate act to intentionally degrade the satellites ability to perform its mission. BENEFIT: The specific techniques employed are potentially applicable to a broad range of time series analysis problems including more general forms of failure detection, stock market analysis, etc.