论文部分内容阅读
针对单传感器在多机动目标跟踪系统中不能很好地处理目标数目变化与突发机动的问题,提出了多传感器多机动目标跟踪的概率假设密度滤波算法.以CPHD滤波算法为理论基础,同时递推概率假设密度(PHD)函数和基数分布,避免了多目标多传感器的数据关联问题.结合自适应当前统计模型,选择3个雷达作为跟踪目标的传感器,相比于单传感器降低了信息的模糊度,提高了可信度.仿真结果比较表明了多传感器CPHD滤波算法在多目标跟踪方面的性能优势.
Aiming at the problem that single sensor can not deal with the change of target number and burst maneuver well in multi-maneuvering target tracking system, a probabilistic hypothesis density filtering algorithm for multi-maneuvering multi-maneuvering target tracking is proposed. Based on CPHD filtering algorithm, Push Probability Hypothesis Density (PHD) function and cardinality distribution, to avoid the data association problem of multi-target multi-sensor.Combining with the current statistical model, three radars are selected as the tracking target sensors, which reduces the information ambiguity compared with the single sensor The reliability of multi-target CPHD filter algorithm is improved by comparison of simulation results.