Multiple Linear Regression to detect shielding effectiveness degradations

Abstract

Multiple linear regression can be successfully used to detect the shielding effectiveness (SE) degradation of enclosures by comparing the results of tests performed at regular time intervals. Simulations are performed on a rectangular metal enclosure with rectangular and circular apertures of various sizes. The frequency curves of the voltages picked by sensors located inside shielded enclosures are interpolated with multiple linear regression. The regression coefficients interpolating the results of the simulations using the same simulation setups are compared; their variations give the indication that the SE has degraded.