摘要
针对基于粒子群优化算法的粒子滤波精度不高,容易陷入局部最优,难以满足目标跟踪的问题,提出了一种新的粒子群优化粒子滤波算法,该算法利用社会个体对群体的认知规律优化了粒子更新的方法,并且完善了粒子速度的更新策略,使优势速度有较小概率变异,从而提高了寻优能力,同时将劣势速度随机初始化,保证了样本的多样性.实验结果表明,该算法精度高,鲁棒性强,可以有效地应用于雷达机动目标跟踪.
Particle filter based on particle swarm optimization algorithm is not precise and easily traps in local optimum, and it is difficult to satisfy the requirement of target tracking. To solve these problems, a novel particle swarm optimized particle filter is proposed. The method for updating particles is optimized by analyzing the cognition rule of individuals to groups, and the speed update strategy is improved. As a result, the superior particle velocity can mutation with a small probability, which improves the search ability. Meanwhile, due to the random evaluation for inferior particle, the diversity of filter is ensured. The simulation results show that this algorithm has the high precision, strong robustness and it＇s suitable for radar target tracking.
出处
《信息与控制》
CSCD
北大核心
2012年第4期413-418,共6页
Information and Control
基金
国防重点预研项目(40405020201)
高等学校博士学科点专项科研基金资助课题(200802881017)
关键词
粒子群优化
粒子滤波
目标跟踪
闪烁噪声
particle swarm optimization
particle filter
target tracking
glint noise
