摘要
针对无线传感器网络实际环境的非线性模型目标跟踪问题,提出一种改进的粒子滤波跟踪算法。首先用模糊C-均值算法确定量测的目标归属,对同一目标的量测进行线性融合,然后用采样重要重采样粒子滤波估计目标位置。仿真结果表明:在非线性模型下,所提出算法与扩展卡尔曼滤波相比,目标估计位置的均方根误差从0.689 5m显著减小到0.370 3m。
Aiming at non- linear model in the physical environment of wireless sensor network, an improved algorithm of particle filter is proposed for target tracking. First, the observed data from the sensors were fuzzy clustered by fuzzy C- means algorithm and integrated by linear fusion, then estimated the location of target by sampling importance resampling particle filter. The results of simulation show that the of target estimated location decreases greatly from 0.689 5 to 0.370 3 compared with extended Kalman filter for the non - linear model.
出处
《茂名学院学报》
2010年第1期37-40,48,共5页
Journal of Maoming College
基金
广东省自然科学基金项目
茂名市重点科技计划项目
关键词
目标跟踪
非线性模型
扩展卡尔曼滤波
粒子滤波
target tracking
nonlinear model
extended kalman filter
particle filtering
