Multi-Hypothesis Tracking Algorithm for Missile Group Targets

This paper addresses the challenge of group targets tracking in complex environments, aiming to enhance the real-time performance and accuracy of tracking systems by optimizing multi-dimensional assignment strategy for selecting the globally optimal hypothesis. A new tracking algorithm named Ballistic Track-Oriented MHT (BTOMHT) is proposed. This algorithm introduces Lagrangian multipliers to achieve effective dimensionality reduction, thereby significantly decreases the computation complexity of data association in multi-dimensional spaces for group targets. Simulations demonstrate that compared with the Global Nearest Neighbor algorithm, BTOMHT not only sustains high precision but also has a great performance in complex dynamic tracking scenarios. It can enhance the efficiency of both target detection and tracking and concurrently bolster overall system stability in tracking tasks. The efficiency of the proposed algorithm is validated across two assessment indices. The algorithm contributes to the theoretical framework of multi-target tracking and furnishes technical support for applications in military defense.