Precession period estimation of ballistic target in midcourse

摘要

To solve the problem of inaccuracy and bad anti-noise performance in traditional methods on precession period estimation, a novel approach is suggested. The approach is based on the exploiting average magnitude difference function (AMDF) and autocorrelation function (ACF). No a priori information about ballistic target under classification is required for estimation of period. The trend term of AMDF is eliminated after empirical mode decomposition (EMD) processes. Then the ACF is used to weight the new AMDF. Thus the new feature is constructed to estimate precession period. Synthetic data of ballistic target is used both for performance evaluation of suggested and traditional approaches. The benefit of suggested approach comparing to traditional approaches is demonstrated using the mean square errors of precession period estimates computed in additive noise environment.