中文摘要：

受遗传算法“杂交”思想的启发,提出一个改进带收缩因子的粒子群算法。该算法在带收缩因子的粒子群算法中增加扰乱粒子认知能力的方法。即对粒子i,随机选择另外一个粒子j,按照一定的概率用粒子j的当前位置替换粒子i的当前位置。为了检验新算法的性能,选用5个高维函数进行了测试,实验结果表明,改进的算法不仅具有良好的稳健性,而且还有良好的收敛性。

英文摘要：

Enlightened by the crossover of genetic algorithm, a novel algorithm of improving panicle swarm optimization with constriction factor by disturbing the cognitive capability of the particle is presented. For each particle i in the swarm, another particle j is selected. The current position of the i-th particle is replaced by the current position of the j-th panicle randomly. Five high-dimensional functions are selected to test the performance of the new algorithm, the result of experiment verified that the proposed algorithm had better stabilization and convergence.