针对标准粒子滤波的粒子贫化问题，提出一种基于改进高斯粒子滤波（Improved Gaussian Particle Filter，IGPF）的船舶非线性状态估计器。首先基于序贯重要性采样（Sequential Important Sampling，SIS）的理论框架，给出标准粒子滤波（Particle Filter，PF）算法，然后在此基础上用高斯分布来作为重要性分布给出高斯粒子滤波（Gaussian Particle Filter，GPF），并将Unscented卡尔曼滤波用于重要性密度函数形成IGPF，应用于动力定位船的状态估计。仿真结果表明，基于IGPF的非线性状态估计器能有效避免粒子贫化、估计船舶状态，并对观测野值有一定的鲁棒性。
Aiming at the problem of particle degen-erace in Standard Particle Filter (SPF), a ships nonlinear state estimate filter based on Improved Gaussian Particle Filter (IGPF) is proposed. Based on the theoretical framework of Sequential Important Sampling, the SPF and its algorithm process is given. The Gaussian Particle filter take Gaussian distribution as importance distribution, we proposed the IGPF which utilize the Unscented Kalman Filter to replace original importance density function and applied it to the state estimate of vessel in the dynamic positioning systems. The simulation results verify the effectiveness of avoiding particle degen-erace of IGPF and robustness of the outliers from measuring sensors.
2016,(s1): 37-43 收稿日期：2016-07-31
DOI：10.3404/j.issn.1672-7619.2016.S1.006
分类号：TP202
作者简介：林孝工(1962-),男,教授,博士生导师,主要研究方向为船舶动力定位系统及多传感器数据融合等。
参考文献：
[1] SØRESEN A J. A survey of dynamic positioning control systems[J]. Annual Reviews in Control, 2011, 35(1):123-136.
[2] 边信黔, 付明玉, 王元慧. 船舶动力定位[M]. 北京:科学出版社, 2011.Bian Xin-qian, FU Ming-yu, WANG Yuan-hui. Dynamic positioning[M]. Beijing:Science Press, 2011.
[3] SUNAHARA Y. An approximate method of state estimation for nonlinear dynamical systems[C]//Joint Automatic Control Conference. Colorado:University of Colorado, 1969.
[4] JULIER S J, UHLMANN J K. New extension of the Kalman filter to nonlinear systems[C]//Proceedings of Aero Sence:the 11th International Symposium on Aerospace/Defense Sensing, Simulation and Controls. Orlando, Florida:SPIE, 1997.
[5] ARASARATNAM I, HAYKIN S. Cubature Kalman filters[J]. IEEE Transactions on Automatic Control, 2009, 54(6):1254-1269.
[6] GILKS I L, BERZUINI C. Following a moving target-Monte Carlo inference for dynamic Bayesian models[J]. Journal of the Royal Statistical Society Series B:Statistical Methodology, 2001, 63(1):127-146.
[7] SHREIDER Y A. Method of statistical testing:Monte Carlo method)[M]. Oxford:Pergamon Press, 1964.
[8] 赵琳. 非线性系统滤波理论[M]. 北京:国防工业出版社, 2012:6-8.ZHAO Lin. Nonlinear system filtering theory[M]. Beijing:National Defense Industry Press, 2012:6-8.
[9] VO B N, SINGH S, DOUCET A. Sequential Monte Carlo methods for multitarget filtering with random finite sets[J]. IEEE Transactions on Aerospace and Electronic Systems, 2005, 41(4):1224-1245.
[10] MOHAMMADI A, ASIF A. Consensus-based distributed unscented particle filter[C]//Proceedings of 2011 IEEE Statistical Signal Processing Workshop (SSP). Nice:IEEE, 2011:237-240.
[11] 曲彦文, 张二华, 杨静宇. 改进的无迹粒子滤波算法[J]. 控制理论与应用, 2010, 27(9):1152-1158.QU Yan-wen, ZHANG Er-hua, YANG Jing-yu. Improved unscented particle filter[J]. Control Theory & Applications, 2010, 27(9):1152-1158.
[12] ZHAO Y X, XING W, YUAN H R, et al. A collaborative control framework with multi-leaders for AUVs based on unscented particle filter[J]. Journal of the Franklin Institute, 2016, 353(3):657-669.
[13] KOTECHA J H, DJURIC P M. Gaussian particle filtering[J]. IEEE Transactions on Signal Processing, 2003, 51(10):2592-2601.
[14] 李殿璞. 船舶运动与建模[M]. 北京:国防工业出版社, 1999:1-17.LI Dian-pu. Ship motion and modeling[M]. Harbin:Harbin Engineering University Press, 1999:1-17.
[15] FOSSEN T I. Marine control systems:guidance, navigation and control of ships, rigs and underwater vehicles[M]. Trondheim, Norway:Marine Cybernetics, 2002.
[16] KLOEDEN P E, PLATEN E. Higher-order implicit strong numerical schemes for stochastic differential equations[J]. Journal of Statistical Physics, 1992, 66(1/2):283-314.
