摘要
针对PSO在计算后期多样性不足、易发生优化停滞的现象,引入免疫系统中的阴性选择概念,定义了新的计算亲和力的方法,提出了带阴性选择的粒子群优化算法,并对其进行了计算复杂性分析.改进算法能在检测到粒子群收敛至局部解后,更新群体中的部分粒子,并使新粒子在解空间上远离局部解,提高了粒子的多样性.试验证明,改进算法的优化性能优于PSO和局部PSO.对改进算法的计算成本及参数选择进行了讨论,并提出了下一步的研究方向.
In consideration of stagnation phenomenon in the later phase of the particle swarm optimization caused by diversity scarcity of particles, the concept of negative selection of the immune system was introduced and a method for affinity computation was defined. On the basis of the above analysis, a particle swarm optimization with negative selection was proposed and its computing complexity was analyzed. The improved algorithm replaced part of the swarm with new particles and made the new ones far from the local optimum in solution space. By means of that, the diversity of particles was promoted. Experiments showed that performance of the algorithm was superior to PSO （Particle Swarm Optimization） and local PSO. The computing cost of the proposed algorithm, its parameters selection and the further research direction were discussed.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2006年第2期87-90,共4页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
关键词
粒子群优化
阴性选择
免疫系统
particle swarm optimization
negative selection
immune system
