Evaluation of Methodologies for Estimating Vulnerability to Electromagnetic Pulse Effects

摘要

Estimation of vulnerability to high-altitude electromagnetic pulse (EMP) effects is essential for strategic and tactical decisions affecting national security. Both the design and the assessment of protection against EMP are inherently subject to uncertainty. The reason is that these processes must be conducted without exposure to actual EMP, in contrast to the situation for other forms of electrical overstress. Estimating vulnerability of systems to EMP effects depends greatly on the nature of the system. The soundest results can be obtained where stress within the system is controlled, through integral shielding and penetration-control devices, to well known values. In this case, one can rely on engineering analysis and systematic testing of a predominantly deterministic nature. Where control and knowledge of stress, as well as of strength, are not possible because of system design, complexity, or uncontrolled changes, probabilistic estimates become necessary. Statistical methods for estimating and combining uncertainties, fault tree analysis, and Bayesian inference may be used to systematize the estimates of vulnerability. However, repeated testing of systems, and subsystems, at as high a simulated threat level as possible, is essential with this approach. Whatever method is used, the uncertainty of the result should be clearly emphasized to decision makers lest oversimplification result.