This paper proposes a particle swarm optimization（PSO） based particle filter（PF） tracking framework,the embedded PSO makes particles move toward the high likelihood area to find the optimal position in the state t...This paper proposes a particle swarm optimization（PSO） based particle filter（PF） tracking framework,the embedded PSO makes particles move toward the high likelihood area to find the optimal position in the state transition stage,and simultaneously incorporates the newest observations into the proposal distribution in the update stage.In the proposed approach,likelihood measure functions involving multiple features are presented to enhance the performance of model fitting.Furthermore,the multi-feature weights are self-adaptively adjusted by a PSO algorithm throughout the tracking process.There are three main contributions.Firstly,the PSO algorithm is fused into the PF framework,which can efficiently alleviate the particles degeneracy phenomenon.Secondly,an effective convergence criterion for the PSO algorithm is explored,which can avoid particles getting stuck in local minima and maintain a greater particle diversity.Finally,a multi-feature weight self-adjusting strategy is proposed,which can significantly improve the tracking robustness and accuracy.Experiments performed on several challenging public video sequences demonstrate that the proposed tracking approach achieves a considerable performance.展开更多
This paper improves the resampling step of particle filtering（PF） based on a broad interactive genetic algorithm to resolve particle degeneration and particle shortage.For target tracking in image processing,this pa...This paper improves the resampling step of particle filtering（PF） based on a broad interactive genetic algorithm to resolve particle degeneration and particle shortage.For target tracking in image processing,this paper uses the information coming from the particles of the previous fame image and new observation data to self-adaptively determine the selecting range of particles in current fame image.The improved selecting operator with jam gene is used to ensure the diversity of particles in mathematics,and the absolute arithmetical crossing operator whose feasible solution space being close about crossing operation,and non-uniform mutation operator is used to capture all kinds of mutation in this paper.The result of simulating experiment shows that the algorithm of this paper has better iterative estimating capability than extended Kalman filtering（EKF）,PF,regularized partide filtering（RPF）,and genetic algorithm（GA）-PF.展开更多
An adaptive particle filter for fault diagnosis of dead-reckoning system was presented,which applied a general framework to integrate rule-based domain knowledge into particle filter.Domain knowledge was exploited to ...An adaptive particle filter for fault diagnosis of dead-reckoning system was presented,which applied a general framework to integrate rule-based domain knowledge into particle filter.Domain knowledge was exploited to constrain the state space to certain subset.The state space was adjusted by setting the transition matrix.Firstly,the monitored mobile robot and its kinematics models,measurement models and fault models were given.Then,5 kinds of planar movement states of the robot were estimated with driving speeds of left and right side.After that,the possible(or detectable)fault modes were obtained to modify the transitional probability.There are two typical advantages of this method,i.e.particles will never be drawn from hopeless area of the state space,and the particle number is reduced.展开更多
对于AGV(Automated Guided Vehicle)传统导航定位方式存在的维护成本高、鲁棒性低,以及原始AMCL(Adaptive Monte Carlo Localization)激光定位方法位姿跟踪精度低、全局定位及机器人“被绑架”问题成功率低的问题,提出了一种基于改进AMC...对于AGV(Automated Guided Vehicle)传统导航定位方式存在的维护成本高、鲁棒性低,以及原始AMCL(Adaptive Monte Carlo Localization)激光定位方法位姿跟踪精度低、全局定位及机器人“被绑架”问题成功率低的问题,提出了一种基于改进AMCL算法的AGV激光定位方法。首先在位姿、权重更新阶段,采用最优粒子滤波器(OPF)取代原始AMCL中用于计算粒子新姿态和权重的传统蒙特卡洛定位方法,提高位姿跟踪精度;其次在重采样阶段,在原始AMCL算法KLD(Kullback-Leibler distance)采样粒子滤波器的基础上,新添加自适应粒子滤波器(SAPF)算法,提高全局定位及机器人“被绑架”问题成功率;最后,针对AGV工作时的特性,在真实车间环境中进行重复定位精度测试,改进后的算法比原算法均方根误差分别提高了54.5%、53.1%、44.7%,且位姿误差可以控制在4 cm、2°以内,满足实际使用要求。展开更多
针对纯方位目标跟踪(Bearing-Only Tracking,BOT)系统强非线性特点,提出一种新的解决方案:采用平方根中心差分卡尔曼滤波器(Square-RootCDKF,SRCDKF)产生粒子滤波提议分布,融入最新的观测数据影响;增加改进措施以提高滤波性能,如采用系...针对纯方位目标跟踪(Bearing-Only Tracking,BOT)系统强非线性特点,提出一种新的解决方案:采用平方根中心差分卡尔曼滤波器(Square-RootCDKF,SRCDKF)产生粒子滤波提议分布,融入最新的观测数据影响;增加改进措施以提高滤波性能,如采用系统重抽样算法减少方差、应用马尔可夫链模特卡罗(Markovchain Monte Carlo,MCMC)方法消除粒子贫乏等。仿真表明该算法是有效的,针对当前BOT系统,比传统EKF、PF算法可靠性更好,跟踪精度更高。展开更多
基金supported by the Chinese Ministry of Science and Intergovernmental Cooperation Project (2009DFA12870)the National Science Foundation of China (60974062,60972119)
文摘This paper proposes a particle swarm optimization（PSO） based particle filter（PF） tracking framework,the embedded PSO makes particles move toward the high likelihood area to find the optimal position in the state transition stage,and simultaneously incorporates the newest observations into the proposal distribution in the update stage.In the proposed approach,likelihood measure functions involving multiple features are presented to enhance the performance of model fitting.Furthermore,the multi-feature weights are self-adaptively adjusted by a PSO algorithm throughout the tracking process.There are three main contributions.Firstly,the PSO algorithm is fused into the PF framework,which can efficiently alleviate the particles degeneracy phenomenon.Secondly,an effective convergence criterion for the PSO algorithm is explored,which can avoid particles getting stuck in local minima and maintain a greater particle diversity.Finally,a multi-feature weight self-adjusting strategy is proposed,which can significantly improve the tracking robustness and accuracy.Experiments performed on several challenging public video sequences demonstrate that the proposed tracking approach achieves a considerable performance.
基金Supported by National Natural Science Foundation of China （60634030 60702066） Aerospace Science Foundation （20090853013） Doctoral Program Foundation of China（20060699032）
基金supported by the National Natural Science Foundation of China(61302145)
文摘This paper improves the resampling step of particle filtering（PF） based on a broad interactive genetic algorithm to resolve particle degeneration and particle shortage.For target tracking in image processing,this paper uses the information coming from the particles of the previous fame image and new observation data to self-adaptively determine the selecting range of particles in current fame image.The improved selecting operator with jam gene is used to ensure the diversity of particles in mathematics,and the absolute arithmetical crossing operator whose feasible solution space being close about crossing operation,and non-uniform mutation operator is used to capture all kinds of mutation in this paper.The result of simulating experiment shows that the algorithm of this paper has better iterative estimating capability than extended Kalman filtering（EKF）,PF,regularized partide filtering（RPF）,and genetic algorithm（GA）-PF.
基金Project(60234030)supported by the National Natural Science Foundation of China
文摘An adaptive particle filter for fault diagnosis of dead-reckoning system was presented,which applied a general framework to integrate rule-based domain knowledge into particle filter.Domain knowledge was exploited to constrain the state space to certain subset.The state space was adjusted by setting the transition matrix.Firstly,the monitored mobile robot and its kinematics models,measurement models and fault models were given.Then,5 kinds of planar movement states of the robot were estimated with driving speeds of left and right side.After that,the possible(or detectable)fault modes were obtained to modify the transitional probability.There are two typical advantages of this method,i.e.particles will never be drawn from hopeless area of the state space,and the particle number is reduced.
文摘对于AGV(Automated Guided Vehicle)传统导航定位方式存在的维护成本高、鲁棒性低,以及原始AMCL(Adaptive Monte Carlo Localization)激光定位方法位姿跟踪精度低、全局定位及机器人“被绑架”问题成功率低的问题,提出了一种基于改进AMCL算法的AGV激光定位方法。首先在位姿、权重更新阶段,采用最优粒子滤波器(OPF)取代原始AMCL中用于计算粒子新姿态和权重的传统蒙特卡洛定位方法,提高位姿跟踪精度;其次在重采样阶段,在原始AMCL算法KLD(Kullback-Leibler distance)采样粒子滤波器的基础上,新添加自适应粒子滤波器(SAPF)算法,提高全局定位及机器人“被绑架”问题成功率;最后,针对AGV工作时的特性,在真实车间环境中进行重复定位精度测试,改进后的算法比原算法均方根误差分别提高了54.5%、53.1%、44.7%,且位姿误差可以控制在4 cm、2°以内,满足实际使用要求。
文摘针对纯方位目标跟踪(Bearing-Only Tracking,BOT)系统强非线性特点,提出一种新的解决方案:采用平方根中心差分卡尔曼滤波器(Square-RootCDKF,SRCDKF)产生粒子滤波提议分布,融入最新的观测数据影响;增加改进措施以提高滤波性能,如采用系统重抽样算法减少方差、应用马尔可夫链模特卡罗(Markovchain Monte Carlo,MCMC)方法消除粒子贫乏等。仿真表明该算法是有效的,针对当前BOT系统,比传统EKF、PF算法可靠性更好,跟踪精度更高。
