论文部分内容阅读
通过两组势阱中心不同且相互协同的主、辅子群,在具有量子行为的粒子群优化(QPSO)算法基础上构造一种基于随机评价机制的交互式双子群QPSO算法(DIR-QPSO).该算法通过子群间的协作避免了种群多样性的快速消失,增强了算法的全局搜索能力.同时,随机因子的加入进一步提高了粒子摆脱局部极值的能力.对6个测试函数的实验结果表明,DIR-QPSO算法相对于传统的粒子群优化算法(PSO)在处理单峰和多峰函数时具有更好的优化性能,收敛速度和收敛精度都得到了较大的提高.
Based on QPSO (Quantum Behaviors Optimization) with quantum behavior, two groups of main and auxiliary subgroups with different and mutually coordinated potential centers are used to construct an interactive bi-subgroup QPSO algorithm (DIR-QPSO) based on random evaluation mechanism. The algorithm avoids the rapid disappearance of population diversity and enhances the global search ability of the algorithm through the collaboration among subgroups.At the same time, the addition of random factors further enhances the ability of particles to get rid of local extremum.Experiments on six test functions The results show that compared with the traditional Particle Swarm Optimization (PSO) algorithm, the DIR-QPSO algorithm has better optimization performance when dealing with single-peak and multi-peak functions, and the convergence speed and convergence precision are greatly improved.