基于混沌变异蝙蝠算法的无人机战场侦察目标跟踪

UAV Object Tracking for Battlefield Reconnaissance Based on a Chaotic-mutated Bat Algorithm

  • 摘要: 为解决粒子滤波重采样过程中的粒子贫化现象,采用了新型启发式算法——混沌变异蝙蝠算法对粒子滤波进行改进,从而实现无人机的目标跟踪.蝙蝠算法是一种基于蝙蝠根据回声频率和响度变换定位机制的群智能启发算法,该方法在许多优化问题中具有良好的性能.在粒子滤波的过程中采用蝙蝠优化算法进行粒子择优,并利用混沌变异策略对蝙蝠算法进行了改进,从而避免后期重采样产生的粒子贫化等问题.本方法用于无人机在战场中进行准确的目标侦察和跟踪,为验证所提出的跟踪策略的有效性,采用传统粒子滤波的跟踪方法与之进行仿真对比实验.实验结果表明,所提方法性能优于其它对比方案且获取的目标函数适应度值高于标准的优化算法.

     

    Abstract: To solve the particle dilution problem of particle filter during the resampling process, we propose a novel unmanned aerial vehicle (UAV) object tracking algorithm based on a new heuristic algorithm called chaotic mutated bat algorithm. The bat algorithm (BA) is inspired by the echolocation mechanism of bats in accordance with pulse rates variance of emission and loudness. It has been proven effective in a wide range of optimization problems. We adopt the BA in UAV object tracking by using particle filter, and a chaotic-mutated improvement is proposed by using chaotic theory and a mutated operator for BA. We were able to resolve the particle dilution problem in the particle filter. To justify the effectiveness of the proposed method, comparative tracking simulations are conducted by employing the standard particle filter. Experimental results show that the proposed method is superior to other methods, which achieves a higher fitness value than standard optimization algorithms.

     

/

返回文章
返回