论文部分内容阅读
为了保持群体多样性以增强全局搜索能力,小生境技术在遗传算法中得到了广泛应用.针对多模态函数优化问题,将小生境技术引入到粒子群算法中,建立小生境熵作为群体多样性的量化指标,实时考查进化过程中群体的多样性并调整进化参数;结合数论中的佳点理论,提出一种在解空间使用佳点搜索的群体多样性发掘方法,使得进化过程中群体多样性水平始终保持在设定的阈值之上,从而改善算法的全局搜索能力以期跳出局部最优;在此基础上提出一种旨在找出全部全局最优解和局部最优解的新型串行多群体小生境粒子群算法.数值实验表明,改进的小生境粒子群算法在求解多模态函数优化问题时具有较好的自适应性和收敛性.将算法应用于图像配准实验中,使得配准参数估计误差有明显降低.
In order to maintain the population diversity and enhance the global search ability, niche technology has been widely used in genetic algorithms.In order to solve the multi-modal function optimization problem, the niche technology is introduced into the PSO, and the niche entropy is established as the population diversity In the meantime, combining with the good point theory in number theory, this paper proposes a method of group diversity discovery using good point search in solution space, which makes the diversity of population in the process of evolution The level is always above the set threshold, so as to improve the global search ability of the algorithm in order to jump out of the local optimum. On the basis of this, a new type of serial multi-objective solution is proposed to find all global optimal solutions and local optimal solutions Population niche PSO algorithm.Numerical experiments show that the improved niche PSO algorithm has better adaptability and convergence in solving multi-modal function optimization problems.The algorithm is applied to image registration experiments, Quasi-parameter estimation error is significantly reduced.