基于IMM-BF的自适应扩张箱粒子机动目标跟踪算法

Adaptive Extended Box Particle Maneuvering Target Tracking Algorithm Based on IMM-BF

针对箱粒子滤波算法在杂波量测环境下跟踪机动目标精度不足和目标丢失的问题,提出一种基于交互多模型伯努利滤波的自适应扩张箱粒子机动目标跟踪(Interacting Multiple Model-Extended Box Particle-Bernoulli Filter, IMM-EBox-BF)算法,采用多个模型并行滤波,在预测步骤后引入自适应箱粒子扩张算法,在每个箱粒子分割成小箱粒子后自适应扩张小箱粒子区间长度,以提高对目标位置估计精度。在更新步骤,改进箱粒子收缩算法,增加对加速度分量的约束,以提高对目标速度估计精度。对仿真与实测数据的处理结果表明,在杂波量测和传感器发生漏检情况下,所提的IMM-EBox-BF算法与传统算法相比,位置跟踪精度提升了16.5%,具备更准确的目标估计精度和连续性。

To solve the problem of insufficient tracking accuracy and target loss of the box particle filter algorithm in the clutter measurement environment,an adaptive extended box particle maneuvering target tracking algorithm based on Interacting Multiple Model-Extended Box Particle-Bernoulli Filter(IMM-EBox-BF)is proposed.The algorithm uses multiple model parallel filtering.After the prediction step,the adaptive box particle expansion algorithm is introduced.After each box particle is divided into small box particles,the interval length of the small box particles is adaptively expanded,which improves the accuracy of target position estimation.In the update step,the box particle contraction algorithm is improved,the constraint on the acceleration component is increased to improve the accuracy of the target velocity estimation.The simulation and measured data processing results show that the proposed IMM-EBox-BF algorithm improves the position tracking accuracy by 16.5%compared with the traditional algorithm,and has more accurate target estimation accuracy and continuity in the case of clutter measurement and missed detection of sensor.