一种基于遗传算法的改进粒子滤波器
|
摘要
针对传统粒子滤波器中粒子权值退化和粒子多样性散失的问题,结合云自适应遗传算法提出一种改进的粒子滤波算法。该算法利用遗传算法中的交叉、变异操作来保证粒子的有效性和多样性。在改进的算法中,引入了云模型自适应遗传算法,利用云滴的随机性和稳定倾向性特点,自适应地调整交叉和变异的概率,既保护了有效粒子不被破坏,又具有随机性,提高了粒子的多样性,从而提高粒子对系统状态变化的适应性。仿真表明,改进算法有效提高了非线性系统状态的估计精度.减少了运行时间。
Specific to the problem of particle weight degradation and loss of diversity in the traditional particle filter, this paper puts forward a modified particle filter algorithm combined with cloud adaptive genetic algorithm. The algorithm uses the crossover and mutation operations in the genetic algorithm to ensure the effectiveness and diversity of the particles. In the improved algorithm, the adaptive genetic algorithm of cloud model was introduced. By using the randomness and stability tendency of cloud droplets, the probability of crossover and mutation was adjusted adaptively,which not only protects the effective particles from being destroyed, but also has randomness to improve the diversity of particles, thereby improving the adaptability of particles on the state changes in the system. The simulation results show that the algorithm proposed in this paper can improve the estimation accuracy of the nonlinear system and reduce the running time.
Specific to the problem of particle weight degradation and loss of diversity in the traditional particle filter, this paper puts forward a modified particle filter algorithm combined with cloud adaptive genetic algorithm. The algorithm uses the crossover and mutation operations in the genetic algorithm to ensure the effectiveness and diversity of the particles. In the improved algorithm, the adaptive genetic algorithm of cloud model was introduced. By using the randomness and stability tendency of cloud droplets, the probability of crossover and mutation was adjusted adaptively,which not only protects the effective particles from being destroyed, but also has randomness to improve the diversity of particles, thereby improving the adaptability of particles on the state changes in the system. The simulation results show that the algorithm proposed in this paper can improve the estimation accuracy of the nonlinear system and reduce the running time.
引文
[1]S Arulampalam,S Maskell,N Gordon.A tutorial on particle filters for online nonlinear/nonGaussian Bayesian tracking[J].IEEE Trans Signal Process,1993,140(2):107-113.
[2]王法胜等.v粒子滤波算法[J].计算机学报,2014,37(8):1679-1694.
[3]P Moral,A Doucet,A Jasra.On adaptive resampling strategies for sequential Monte Carlo methods[J].Bernoulli,2012,18(1):252-278.
[4]J P Zhong,Y F Fung.A biologically inspired improvementstategy for particle filter:An tcolony optimization assisted particle filter[J].International Journal of Control,Automation and System,2010,8(3):519-526.
[5]吴宝成.粒子滤波重采样算法研究及其应用[D].哈尔滨工业大学计算机科学与技术学院,2006.
[6]冯驰,赵娜,王萌.一种改进残差重采样算法的研究[J].哈尔滨工程大学学报,2010,31(1):120-124.
[7]SebastianThrun,Wolfram Burgard,Dieter Fox[M].Probabilistic Robotics,2002,The MIT Press,2002.
[8]A Doucet,N Gordon.Sequentia 1 monte Ca rlo methodsin practice[M].New Yor k:Springer-Verlag,2001.
[9]Aratyn,Henrik,Rasinariu,Constantin.A short course in mathematical methods with Maple[J].World Scientific,2006.
[10]张铃,张钹.遗传算法机理的研究[J].软件学报,2000,11(7):945-952.
[11]戴朝华,等.云自适应遗传算法[J].控制理论与应用,2007,24(4):646-651.
[12]Hollandjh.Adaptation in Natural and Artificial System:An Introductory Analysis with Applications to Biology,Control and Artificial Intelligence(2nded)[M].Cambridge:MIT Press,1992.
[13]刘莉,陈学允.基于模糊遗传算法的配电网络重构[J].中国电机工程学报,2000,20(2):66-69.
[14]刘学增,周敏.改进的自适应遗传算法及其工程应用[J].同济大学学报,2009,37(3):303-307.
[15]C Y Li,H B Ji.A new particle filter with GA-MCMC resampling[A].Proceedings oflnternational Conference on Wavelet Analysis and Patternrecognition[C].Beijing,China,2007:146-150.
[16]X Fu,Y Jia.An improvement on resampling algorithm ofparticle filters[J].IEEE Transactions on Signal Processing,2010,58(10):5414-5420.
[17]李德毅,刘常显.论正态云模型的普适性[J].中国工程科学,2004,6(8):28-34.
[18]冯东青,等.遗传算法中选择交叉策略的改进[J].计算机工程,2008,34(19):189-194.
[19]王永良.空间谱估计理论与算法[M],北京:清华大学出版社,2004.
[20]N JGordon,D J Salmond,A F Smith.Novel approach to nonlinear non-Gaussian Bayesian state estimation[J].IEEE Proceedings on radar.Sonar and Navigation,1993,140(2):107-113.
[2]王法胜等.v粒子滤波算法[J].计算机学报,2014,37(8):1679-1694.
[3]P Moral,A Doucet,A Jasra.On adaptive resampling strategies for sequential Monte Carlo methods[J].Bernoulli,2012,18(1):252-278.
[4]J P Zhong,Y F Fung.A biologically inspired improvementstategy for particle filter:An tcolony optimization assisted particle filter[J].International Journal of Control,Automation and System,2010,8(3):519-526.
[5]吴宝成.粒子滤波重采样算法研究及其应用[D].哈尔滨工业大学计算机科学与技术学院,2006.
[6]冯驰,赵娜,王萌.一种改进残差重采样算法的研究[J].哈尔滨工程大学学报,2010,31(1):120-124.
[7]SebastianThrun,Wolfram Burgard,Dieter Fox[M].Probabilistic Robotics,2002,The MIT Press,2002.
[8]A Doucet,N Gordon.Sequentia 1 monte Ca rlo methodsin practice[M].New Yor k:Springer-Verlag,2001.
[9]Aratyn,Henrik,Rasinariu,Constantin.A short course in mathematical methods with Maple[J].World Scientific,2006.
[10]张铃,张钹.遗传算法机理的研究[J].软件学报,2000,11(7):945-952.
[11]戴朝华,等.云自适应遗传算法[J].控制理论与应用,2007,24(4):646-651.
[12]Hollandjh.Adaptation in Natural and Artificial System:An Introductory Analysis with Applications to Biology,Control and Artificial Intelligence(2nded)[M].Cambridge:MIT Press,1992.
[13]刘莉,陈学允.基于模糊遗传算法的配电网络重构[J].中国电机工程学报,2000,20(2):66-69.
[14]刘学增,周敏.改进的自适应遗传算法及其工程应用[J].同济大学学报,2009,37(3):303-307.
[15]C Y Li,H B Ji.A new particle filter with GA-MCMC resampling[A].Proceedings oflnternational Conference on Wavelet Analysis and Patternrecognition[C].Beijing,China,2007:146-150.
[16]X Fu,Y Jia.An improvement on resampling algorithm ofparticle filters[J].IEEE Transactions on Signal Processing,2010,58(10):5414-5420.
[17]李德毅,刘常显.论正态云模型的普适性[J].中国工程科学,2004,6(8):28-34.
[18]冯东青,等.遗传算法中选择交叉策略的改进[J].计算机工程,2008,34(19):189-194.
[19]王永良.空间谱估计理论与算法[M],北京:清华大学出版社,2004.
[20]N JGordon,D J Salmond,A F Smith.Novel approach to nonlinear non-Gaussian Bayesian state estimation[J].IEEE Proceedings on radar.Sonar and Navigation,1993,140(2):107-113.