基于交互多模型的粒子滤波导引头机动目标检测技术研究

Research on Maneuvering Target Tracking Technology for Seeker Based on Particle Filter of Interacting Multiple Model

随着武器装备技术的发展,有人机、无人机等空中飞行器的机动能力大幅提升。目标机动时,由于目标过载未知,其与导引头预置模型间的匹配性下降,通常会导致导引头跟踪性能下降,甚至丢失目标。面对传统导引头目标跟踪算法对大机动目标适应性较差问题,本文提出了一种基于距离-多普勒二维谱粒子滤波的交互多模型算法,通过交互多模型引入多模态,实时进行机动参数辨识,提高模型匹配性,结合粒子滤波弱小信号探测方面的优势提升导引头对大机动目标的适应性。仿真结果表明:基于交互多模型的粒子滤波算法明显改善了雷达导引头对大机动目标的检测跟踪性能,改善了传统的粒子滤波算法对大过载机动的适应性,具有较好工程应用价值。

With the development of weapon equipment technology,the maneuverability of aerial vehicles such as manned aircraft and UAV has been greatly improved.When the target is maneuvered,the matching between the target and the preset model of the seeker is degraded due to the unknown target overload,which usually leads to the degradation of the seeker tracking performance and even the loss of the target.Faced with the problem that the traditional seeker target tracking algorithm has poor adaptability to large maneuvering targets,this paper proposes an interacting multiple model algorithm based on particle-filtering of distance-Doppler two-dimensional spectrum.Led the multi-modality into the interacting multiple model,the maneuvering parameter identification is improved in real time to improve the model match-ing,and the advantage of the small-signal detection of the particle filter technology is combined to improve the adaptability of the seeker to the high maneuvering targets.Simulation results validate that the proposed PF algorithm based on interacting multiple model improves detection and tracking performance of the radar seeker to high maneuvering targets significantly.In addition,it improves the adaptability of PF algorithm to mobile overload and has more important engineering application value.