无序量测下的机动弱目标检测前跟踪

Track-before-detect with Out-of-sequence Measurement for Maneuvering Weak Target

针对无序量测(OOSM)情况下的机动微弱目标检测与跟踪问题,提出一种基于OOSM和多模粒子滤波(OOSM-MMPF)的检测前跟踪(TBD)算法。该算法通过直接利用OOSM对粒子权重进行更新,并在此基础上对粒子集进行重采样,从而实现OOSM情况下的目标状态更新。由于充分利用了OOSM包含的信息,该算法可以有效提高机动微弱目标的正确检测概率与跟踪精度。仿真结果表明,该算法可以有效处理OOSM问题,实现对机动弱目标的有效检测和跟踪,其算法性能接近顺序量测滤波时的MMPF算法性能。

To address the problem of maneuvering weak target detection and tracking with out-ofsequence measurements（OOSM）,a maneuvering weak target track-before-detect method is proposed based on the OOSM and multiple model particle filter（OOSM-MMPF）. By updating the weights of particles directly with the OOSM and performing a resampling step on the particle set,the target state is updated with the OOSM. Due to make full use of the information contained in the OOSM,the proposed method can improve the correct detection probability and tracking performance of target efficiently.Simulation results demonstrate that the proposed method can solve the problem of OOSM and detect and track maneuvering weak target effectively with the performance of target tracking is close to that of MMPF with in-sequence measurements.