摘要
为了解决杂波环境下多机动目标的数据关联问题,提出一种将粒子滤波器(PF)和模糊数据关联(FDA)相结合的数据关联算法。本算法首先应用粒子滤波方法对目标的状态进行采样,得到样本(粒子),然后结合量测,用模糊数据关联算法分别对每个粒子求得其对各个目标的隶属度,比较各个隶属度的大小,把最大的隶属度作为粒子在滤波中的权值,从而实现多目标的跟踪。另外,在应用FDA的过程中对其中的常量m进行取值分析,讨论了其不同取值对跟踪效果的影响。仿真结果表明,与其他的常规跟踪方法粒子滤波及扩展卡尔曼滤波相比,该方法具有较好的跟踪效果;与粒子滤波结合联合数据关联方法相比,跟踪效果相差不大,但是具有更好的实时性。
To solve the problem of how to fuse the data of multiple maneuvering targets in environment with clutter, an algorithm of data fusion combining Particle Filter （PF） with Fuzzy Data Association （FDA） was proposed. First, PF was used for sampling the state of targets to get the particles. And then FDA was used for obtaining the membership degree of each particle to each target. The maximum membership degree was taken as the weight of the particle in filtering for realizing multi-target tracking. And the efficiency of the constant m of FDA was analyzed. Simulation results showed that the algorithm has better tracking results comparing with PF and Extended Kalman Filter （EKF） , and has better real-time performance comparing with PF associated with JPDA.
出处
《电光与控制》
北大核心
2010年第4期34-37,F0003,共5页
Electronics Optics & Control
关键词
多目标跟踪
数据关联
粒子滤波器
机动目标
multi-target tracking
data association
Particle Filter （PF）
maneuvering target
