A Study on Target/Non-target Identification and Projectile Classification Algorithms Using Counter-battery Radar Measurements

This study presents a two-stage algorithmic framework for target/non-target identification and projectile classification using measurement data from counter-battery radar systems. The first stage focuses on distinguishing ballistic targets—characterized by their predictable trajectories—from non-targets exhibiting autonomous or irregular flight patterns. A trajectory-fitting approach is applied to the radar measurements to evaluate their consistency with ballistic motion models, enabling robust discrimination between targets and non-targets. Once a ballistic target is successfully identified, the second stage estimates its trajectory using a modified data-fitting method that minimizes the residuals of both the measurement and dynamic models. The target's motion is approximated by a polynomial function of time, and the resulting trajectory is used to infer the projectile type. Classification is then performed based on an estimated dynamic parameter, composed of mass, reference area, and drag coefficients, derived from the fitted trajectory. Both identification and classification algorithms rely on data-fitting techniques that capture the underlying dynamics and extract key parameters inherent in counter-battery radar observations. The effectiveness of the proposed method is verified through numerical simulation. The proposed method is expected to enhance the radar system’s ability to autonomously filter threats and classify incoming projectiles with minimal prior information.