一种新的可变数目机动目标联合检测与跟踪算法

来源期刊:中南大学学报(自然科学版)2016年第10期

论文作者:胡秀华 郭雷 李晖晖 燕攀登

文章页码:3424 - 3436

关键词:多模粒子滤波;检测与跟踪;关联判定;目标;混合采样

Key words:multi-model particle filter; detection and tracking; associated determination; target; composite sampling

摘    要:在低信噪比多传感器观测环境下,针对机动目标数目变化时跟踪性能不高的问题,提出一种新的目标联合检测与跟踪算法。依据粒子存在变量进行预测状态粒子集的采样,考虑状态粒子集与当前观测值的关联程度,利用模糊拍卖算法与粒子群优化理论解决状态集与观测集之间的关联问题,并给出目标出现与消失的判定准则,实现粒子权值更新;依据混合采样方法得到包含模型信息和状态信息的粒子集,并按照目标模型概率进行粒子状态融合得到目标局部状态后验估计值和均方误差;最后,对关联的各传感器局部航迹信息进行加权融合,得到各目标的全局状态估计值,并与经典多模粒子滤波算法的仿真试验对比分析。研究结果表明:新算法在运动模型概率估计、状态估计及目标数目估计方面具有有效性。

Abstract: Aiming at the problem of low target tracking performance with the variable number of maneuvering target, a new joint detection and tracking algorithm was put forward for the environment of low signal-to-noise ratio multi-sensor observation. The sampling of predict state particle set was completed according to the existence of the particle. And then, taking the association degree of the sets of particle state and the current observations into account, and using the theory of fuzzy auction algorithm and particle swarm optimization, the association problem between the state and observation sets was solved, and the criteria of target appear and disappear was given, and the updating of particle weight was realized. By the means of composite sampling, the sample particle sets with the model information and status information was obtained, and the local posteriori estimate and the mean square error of target state were given with the target model probability through particle state fusion. Finally, with the weighted fusion of each associated local sensor tracking information, the global state estimation of each target was obtained and compared with the simulation experiment with classical multiple model particle filter algorithm. The results show that the new algorithm is effective in motion model probability estimation, state estimation, and the target number estimation.

相关论文

  • 暂无!

相关知识点

  • 暂无!

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号