Data-driven capturability analysis for pure proportional navigation guidance considering target maneuver

Abstract

A data-driven statistical analysis of the missile’s capture region is performed. The capture region is the region of the initial geometric configuration for pursuer missile against a target in which the missile can intercept the target while satisfying specific constraints. The statistical verification approach has advantages over the analytic approach in that it can deal with various guidance algorithms and target maneuver utilizing numerical simulator. In this study, the verification model is constructed using the Gaussian process regression model. The verification model computes the probability distribution of the target capture over the initial configuration space. The data-driven capturability analysis is conducted for the maneuvering target using the Gaussian process regression model. The capture region derived from the statistical model is compared with the analytic model, and the effectiveness of the active sampling algorithm is demonstrated.