一种基于小生境技术的群智能粒子滤波算法

• 作 者: 刘云龙 林宝军
• 机 构: 中国科学院光电研究院,中国科学院研究生院,
• 出 处: 控制与决策
• 发表日期: 2010年02期
• 关键词: 滤波算法 小生境技术 群智能 粒子群优化 粒子滤波 寻优能力 重采样 观测噪声 似然 fitness

针对基本粒子滤波存在严重的退化问题和重采样技术导致粒子枯竭的问题,提出一种新型粒子滤波算法——基于小生境技术的群智能优化粒子滤波算法.通过多模寻优增强粒子的多样性和寻优能力,使得采样后的粒子向高似然区域移动,从而有效地提高了系统状态估计精度.仿真实验表明,该算法是有效而稳定的.

Aiming at the serious degradation problem of basic particle filter and the problem of particle resurfacing caused by resampling technique, a new particle filter algorithm based on niche optimization is proposed, which is based on the niche technology of group intelligence to improve the particle diversity And the optimization ability to make the sampled particles move to the high-likelihood region, which effectively improves the system state estimation accuracy.The simulation results show that the algorithm is effective and stable.