Abstract:

For the problem that the cuckoo algorithm is easily limited to the local optimum, the ability of the cuckoo algorithm to local optimization and global optimization is balanced by improving the search step value [α] of the cuckoo algorithm and the probability [pα] of the species of an exotic bird in this paper. Combined with particle filtering the improved cuckoo algorithm replaces the resampling process of particle filtering to solve the problem of particle depletion and low estimation accuracy. The experimental results show that the particle diversity of the improved cuckoo optimized particle filter algorithm is improved, which ensures the estimation accuracy.

Key words: particle filtering, particle depletion, cuckoo algorithm, resampling

摘要：

针对布谷鸟算法易限于局部最优的问题，通过对布谷鸟算法的搜索步长值[α]和发现外来鸟卵的物种的概率[pα]进行改进，来平衡布谷鸟算法局部寻优与全局寻优的能力。改进的布谷鸟算法与粒子滤波结合，代替粒子滤波的重采样过程，解决粒子贫化和估计精度低的问题。实验结果表明，改进的布谷鸟优化粒子滤波算法中，粒子的多样性提高，从而保证了估计精度的提高。

关键词: 粒子滤波, 粒子贫化, 布谷鸟算法, 重采样