一种新的自适应粒子滤波单通道盲分离算法
|
摘要
目前,解决成对载波多址单通道盲分离问题的主要方法之一是粒子滤波。以往的分离算法中,粒子数往往是固定的。盲分离粒子滤波算法在经过若干次迭代和重采样过后,存在一些权重数量级非常小的粒子,这些粒子不仅对后验概率密度的贡献甚微,而且会浪费大量的运算时间,导致算法效率低下。为提高效率,根据粒子滤波盲分离的特点,在参数大致收敛之后,采用一种自适应的算法降低粒子数目。此方法在保证了精度的同时,降低了计算复杂度。仿真结果表明,改进的算法相比传统粒子滤波算法复杂度降低了约1/6左右,低信噪比条件下精度比传统算法更高。
At present,particle filter is one of the main methods to solve the single channel blind separation problem of paired carrier multiple access. In previous separation algorithms,the number of particles was usually fixed. In the blind separation particle filter algorithm,after several iterations and resampling,there were some particles with very small magnitude of weight. These particles contributed little to the posterior probability density,and a large amount of computation time would be wasted on these useless particles,resulting in inefficient algorithm. In order to improve the efficiency,an adaptive algorithm was adopted to reduce the number of particles after the parameters were approximately converged according to the characteristics of blind separation of particle filter. The simulation results show that the complexity of the improved algorithm is reduced by 1/6 compared with the traditional particle filter algorithm,and the accuracy of the algorithm is higher than that of the traditional algorithm under low signal noise ratio condition.
At present,particle filter is one of the main methods to solve the single channel blind separation problem of paired carrier multiple access. In previous separation algorithms,the number of particles was usually fixed. In the blind separation particle filter algorithm,after several iterations and resampling,there were some particles with very small magnitude of weight. These particles contributed little to the posterior probability density,and a large amount of computation time would be wasted on these useless particles,resulting in inefficient algorithm. In order to improve the efficiency,an adaptive algorithm was adopted to reduce the number of particles after the parameters were approximately converged according to the characteristics of blind separation of particle filter. The simulation results show that the complexity of the improved algorithm is reduced by 1/6 compared with the traditional particle filter algorithm,and the accuracy of the algorithm is higher than that of the traditional algorithm under low signal noise ratio condition.
引文
1 Chang C,Ansari R.Kernel particle filter for visual tracking.IEEESignal Processing Letters,2005;12(3):242-245
2杨智,严华.基于SVR和粒子滤波的室内导航方法.计算机测量与控制,2016;24(9):231-233Yang Zhi,Yan Hua.Indoor navigation method based on SVR and particle filter.Computer Measurement and Control,2016;24(9):231-233
3程水英,张剑云.裂变自举粒子滤波.电子学报,2008;36(3):500-504Cheng Shuiying,Zhang Jianyun.Fission bootstrap particle filter.Electronic Journal,2008;36(3):500-504
4左军毅,张怡哲,梁彦.自适应不完全重采样粒子滤波器.自动化学报,2012;38(4):647-652Zuo Junyi,Zhang Yizhe,Liang Yan.Adaptive incomplete resampling particle filter.Journal of Automation,2012;38(4):647-652
5赵丰,汤磊,张武,等.一种高实时性粒子滤波重采样算法.系统仿真学报,2009;21(18):5789-5793Zhao Feng,Tang Lei,Zhang Wu,et al.A high real time particle filter resampling algorithm.Journal of System Simulation,2009;21(18):5789-5793
6叶龙,王京玲,张勤.遗传重采样粒子滤波器.自动化学报,2007;33(8):885-887Ye Long,Wang Jingling,Zhang Qin.Genetic resampling particle filter.Journal of Automation,2007;33(8):885-887
7高怡,高雅,高社生.基于似然分布的样本数自适应UPF算法.中国惯性技术学报,2015;23(5):648-652Gao Yi,Gao Ya,Gao Shesheng.Sample number adaptive UPF algorithm based on likelihood distribution.Chinese Journal of Inertial Technology,2015;23(5):648-652
8 Arulampalam M S,Maskell S,Gordon N,et al.A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking.IEEETransactions on Signal Processing,2002;50(2):174-188
9段琢华,蔡自兴,于金霞.移动机器人软故障检测与补偿的自适应粒子滤波算法.中国科学:技术科学,2008;38(4):565-578Duan Zhuohua,Cai Zixing,Yu Jinxia.Adaptive particle filter algorithm for soft fault detection and compensation of mobile robot.Chinese Science:Technical Science,2008;38(4):565-578
10王书朋,姬红兵.用于目标跟踪的自适应粒子滤波算法.系统仿真学报,2010;22(3):630-633Wang Shupeng,Ji Hongbing.Adaptive particle filter algorithm for target tracking.Journal of System Simulation,2010;22(3):630-633
11赵洪宇,王伟,蔡爱华.基于自适应粒子滤波算法的对海跟踪研究.现代雷达,2014;36(3):49-52Zhao Hongyu,Wang Wei,Cai Aihua.Research on sea tracking based on adaptive particle filter algorithm.Modern Radar,2014;36(3):49-52
12孟勃.基于自适应粒子滤波的目标跟踪算法.光子学报,2008;37(增刊1):252-255Meng Bo.Object tracking algorithm based on adaptive particle filter.Photonics Journal,2008;37(S1):252-255
13刘国连.成对载波多址接入信号盲分离技术研究.成都:电子科技大学,2015Liu Guolian.Study on blind separation technology of paired carrier multiple access signals.Chengdu:University of Electronic Science and Technology of China,2015
14万坚,涂世龙,廖灿辉,等.通信混合信号盲分离理论与技术.北京:国防工业出版社,2012:240-242Wan Jian,Tu Shilong,Liao Canhui,et al.Blind separation theory and technology for communication mixed signals.Beijing:National Defense Industry Press,2012:240-242
2杨智,严华.基于SVR和粒子滤波的室内导航方法.计算机测量与控制,2016;24(9):231-233Yang Zhi,Yan Hua.Indoor navigation method based on SVR and particle filter.Computer Measurement and Control,2016;24(9):231-233
3程水英,张剑云.裂变自举粒子滤波.电子学报,2008;36(3):500-504Cheng Shuiying,Zhang Jianyun.Fission bootstrap particle filter.Electronic Journal,2008;36(3):500-504
4左军毅,张怡哲,梁彦.自适应不完全重采样粒子滤波器.自动化学报,2012;38(4):647-652Zuo Junyi,Zhang Yizhe,Liang Yan.Adaptive incomplete resampling particle filter.Journal of Automation,2012;38(4):647-652
5赵丰,汤磊,张武,等.一种高实时性粒子滤波重采样算法.系统仿真学报,2009;21(18):5789-5793Zhao Feng,Tang Lei,Zhang Wu,et al.A high real time particle filter resampling algorithm.Journal of System Simulation,2009;21(18):5789-5793
6叶龙,王京玲,张勤.遗传重采样粒子滤波器.自动化学报,2007;33(8):885-887Ye Long,Wang Jingling,Zhang Qin.Genetic resampling particle filter.Journal of Automation,2007;33(8):885-887
7高怡,高雅,高社生.基于似然分布的样本数自适应UPF算法.中国惯性技术学报,2015;23(5):648-652Gao Yi,Gao Ya,Gao Shesheng.Sample number adaptive UPF algorithm based on likelihood distribution.Chinese Journal of Inertial Technology,2015;23(5):648-652
8 Arulampalam M S,Maskell S,Gordon N,et al.A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking.IEEETransactions on Signal Processing,2002;50(2):174-188
9段琢华,蔡自兴,于金霞.移动机器人软故障检测与补偿的自适应粒子滤波算法.中国科学:技术科学,2008;38(4):565-578Duan Zhuohua,Cai Zixing,Yu Jinxia.Adaptive particle filter algorithm for soft fault detection and compensation of mobile robot.Chinese Science:Technical Science,2008;38(4):565-578
10王书朋,姬红兵.用于目标跟踪的自适应粒子滤波算法.系统仿真学报,2010;22(3):630-633Wang Shupeng,Ji Hongbing.Adaptive particle filter algorithm for target tracking.Journal of System Simulation,2010;22(3):630-633
11赵洪宇,王伟,蔡爱华.基于自适应粒子滤波算法的对海跟踪研究.现代雷达,2014;36(3):49-52Zhao Hongyu,Wang Wei,Cai Aihua.Research on sea tracking based on adaptive particle filter algorithm.Modern Radar,2014;36(3):49-52
12孟勃.基于自适应粒子滤波的目标跟踪算法.光子学报,2008;37(增刊1):252-255Meng Bo.Object tracking algorithm based on adaptive particle filter.Photonics Journal,2008;37(S1):252-255
13刘国连.成对载波多址接入信号盲分离技术研究.成都:电子科技大学,2015Liu Guolian.Study on blind separation technology of paired carrier multiple access signals.Chengdu:University of Electronic Science and Technology of China,2015
14万坚,涂世龙,廖灿辉,等.通信混合信号盲分离理论与技术.北京:国防工业出版社,2012:240-242Wan Jian,Tu Shilong,Liao Canhui,et al.Blind separation theory and technology for communication mixed signals.Beijing:National Defense Industry Press,2012:240-242