Rapid Missile System Characterization Using Computational Intelligence

Abstract

There exist many physical relationships between dynamic systems that humans can infer data from where an exact science may not be possible or plausible to apply. The imprecision or inaccuracy of attempted models and sensors often lead to errors when seemingly simple tasks are left to a machine; a seasoned expert may know from experience how to react under certain conditions for which a robot may not have been trained or may not have sensors to a high enough precision. Fuzzy Logic and Artificial Neural Networks are an attempt to remedy that. If machine algorithms are able to infer solutions from previous knowledge, then new potential can be unlocked in computational applications. The study performed discusses the use of computational intelligence methods in the rapid classification of known and unknown missile systems. When a hostile missile is launched, it is crucial to know the geometry and capabilities of the impending ballistic vehicle. Computational intelligence mechanisms could play a large role in this defense application as well as a complementary tool to existing defense systems.