基于CPHD的PD雷达多目标检测前跟踪方法

A multi-target TBD algorithm based on CPHD for PD radar

针对强杂波背景下传统多目标检测前跟踪方法目标数目估计误差大、状态估计精度严重下降的问题,基于势化概率假设密度(CPHD)滤波理论提出一种多普勒雷达多目标检测前跟踪(TBD)方法.该方法在建立脉冲多普勒雷达多目标TBD观测模型的基础上,通过推导基于多普勒量测及具有幅度信息的扩维CPHD预测和更新方程,对目标状态和数目进行估计.在CPHD更新时,首先对目标位置更新,然后序贯使用多普勒量测进行二次更新,并采用高斯混合(Gaussian mixture)近似方法实现,提高估计精度和准确性.仿真结果表明,该方法可有效抑制杂波,实现对多个目标的跟踪.

Aiming at the problems that in strong clutter backgrounds traditional multi-target track-before-detect(TBD) method leads to a great estimation error on target number and a sharp decline of state estimation accuracy, this paper proposes a multi-target TBD algorithm for PD radar based on the cardinalized probability hypothesis density(CPHD) theory. In this method, the TBD observation model for PD radar is built firstly. Secondly, the prediction equation and update equation of an augmented dimension CPHD based on both Doppler measurements and amplitude information are deduced in order to estimate the targets' state and number.In updating CPHD, target location update is done before being updated sequentially by using Doppler measurement and Gaussian mixture approximation is adopted to realize more accuracy and better veracity of estimation. Simulation results show that the proposed algorithm is able to effectively suppress clutter so as to track multiple targets.