摘要
针对交互式多模型粒子滤波在跟踪机动目标时精度受限问题,提出一种基于交互式多模型(IMM)的多传感器顺序粒子滤波算法。采用IMM机制实现目标运动模式的确认;在合理利用单传感器量测和多传感器量测中冗余和互补信息的基础上,引入顺序重抽样方法改善粒子分布,并将改善后的粒子应用于IMM粒子滤波算法框架。仿真实验结果表明:新算法能够估计出强机动目标状态,且精度明显优于标准IMM粒子滤波算法。
Aiming at the precision of the interacting multiple models（IMM） with particle filtering is restricted in tracking maneuvering target,a new IMM which adopts the multi-sensor sequential particle filtering algorithm is proposed.In the novel algorithm,the IMM mechanism is used to confirm the motion pattern of target.Based on utilization of the redundancy and complementary measurement information from the single sensor and multi-sensor,a sequential re-sampling method is used to optimize the distribution of particles and integrated into the IMM with particle filtering.Simulation results show that the new algorithm can estimate the state of the maneuvering target well and is superior to the standard IMM with particle filtering algorithm in the precision.
出处
《传感器与微系统》
CSCD
北大核心
2012年第4期130-132,136,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(60974062)
河南省基础与前沿研究计划资助项目(112300410117)
河南省高等学校青年骨干教师计划资助项目(2010GGJS-041)
关键词
多源信息融合
机动目标跟踪
交互式多模型
粒子滤波
multi-source information fusion
maneuvering target tracking
interacting multiple models（IMM）
particle filtering
