杂波环境下基于全邻模糊聚类的联合概率数据互联算法

针对杂波环境下的多目标跟踪数据互联问题，该文提出基于全邻模糊聚类的联合概率数据互联算法(Joint Probabilistic Data Association algorithm based on All-Neighbor Fuzzy Clustering, ANFCJPDA)。该算法根据确认区域中量测的分布和点迹-航迹关联规则构造统计距离，以各目标的预测位置为聚类中心，利用模糊聚类方法，计算相关波门内候选量测与不同目标互联的概率，通过概率加权融合对各目标状态与协方差进行更新。仿真分析表明，与经典的联合概率数据互联算法(Joint Probabilistic Data Association algorithm, JPDA)相比，ANFCJPDA较大程度地改善了算法的实时性，并且跟踪精度与JPDA相当。

Joint Probabilistic Data Association Algorithm Based on All-neighbor Fuzzy Clustering in Clutter

This paper proposes a new Joint Probabilistic Data Association algorithm based on All-Neighbor Fuzzy Clustering (ANFCJPDA) for mutitarget tracking in the clutter. Firstly, distance measure is established according to measurements distribution in validation area and data correlation rules. Then, the predicted position is set up as a cluster center, and the association probabilities are calculated on the basis of fuzzy clustering, which are used as weights to update targets state and the covariance. Simulation results show that the proposed method reduces highly the computational complexity compared to conventional Joint Probabilistic Data Association (JPDA) technique, and is effective for multiple target tracking in a cluttered environment.