摘要： 将状态转移先验分布和观测似然分布相结合，提出一种基于自适应退火参数优化混合建议分布的粒子滤波算法。根据当前最新的观测信息，以退火参数因子调控混合建议分布中状态转移先验分布与似然建议分布的混合率。在混合建议分布中结合自适应参数优化机制动态调整退火参数。仿真实验验证了该算法的有效性。
关键词:
粒子滤波,
混合建议分布,
状态转移先验分布,
似然建议分布,
退火参数,
自适应参数优化
Abstract: This paper proposes a particle filtering algorithm based on hybrid proposal distribution of adaptive optimization. Based on performance analysis of particle filter with different proposal distribution, hybrid proposal distribution is used to consider the current latest observational information. Annealing parameter factor is utilized to adjust the mix ratio of state transition prior distribution and likelihood proposal distribution in hybrid proposal distribution. By the analysis of hybrid proposal distribution with fixed annealing parameter optimization, adaptive parameter optimization mechanism is combined in above hybrid proposal distribution to dynamically adjust the annealing parameter. With the simulation program of object tracking, the performance of proposed particle filtering algorithm is evaluated and its validity is verified.
Key words:
particle filtering,
hybrid proposal distribution,
state transition prior distribution,
likelihood proposal distribution,
annealing parameter,
adaptive parameter optimization
中图分类号:
于金霞, 汤永利, 许景民, 赵倩. 基于自适应优化混合建议分布的粒子滤波算法[J]. 计算机工程, 2012, 38(16): 200-202.
XU Jin-Xia, SHANG Yong-Li, HU Jing-Min, DIAO Qian. Particle Filtering Algorithm Based on Hybrid Proposal Distribution of Adaptive Optimization[J]. Computer Engineering, 2012, 38(16): 200-202.