靳雁霞, 张鑫, 薛丹. 具有自适应逃逸的环状全互连结构粒子群算法[J]. 微电子学与计算机, 2018, 35(2): 1-5, 10.
引用本文: 靳雁霞, 张鑫, 薛丹. 具有自适应逃逸的环状全互连结构粒子群算法[J]. 微电子学与计算机, 2018, 35(2): 1-5, 10.
JIN Yan-xia, ZHANG Xin, XUE Dan. A Particle Swarm Optimization Algorithm with An Adaptive Escape for A Ring Full Interconnection Structure[J]. Microelectronics & Computer, 2018, 35(2): 1-5, 10.
Citation: JIN Yan-xia, ZHANG Xin, XUE Dan. A Particle Swarm Optimization Algorithm with An Adaptive Escape for A Ring Full Interconnection Structure[J]. Microelectronics & Computer, 2018, 35(2): 1-5, 10.

具有自适应逃逸的环状全互连结构粒子群算法

A Particle Swarm Optimization Algorithm with An Adaptive Escape for A Ring Full Interconnection Structure

  • 摘要: 为了提升粒子群算法在解空间中的寻优性能, 提出了一种具有自适应逃逸机制的基于环状全互连拓扑结构的粒子群算法(RSEPSO).该算法首先将种群中的粒子组成环状结构后再连接成全互连结构, 其次加入自适应逃逸功能, 为防止算法进化时陷入局部极值, 同时选取适应值差的粒子进行重新分布, 并融入优质粒子环作为学习对象, 最后产生足够多的点进行重新搜索, 获得全局最优值.通过4个标准测试函数优化, 与其他优化算法比较, 可以看出RSEPSO能够明显的提升粒子群算法的寻优性能.

     

    Abstract: In order to improve particle swarm optimization algorithm in the solution space to find excellent performance, put forward a with adaptive escape mechanism based on ring interconnection topology of particle swarm optimization algorithm (RSEPSO). In the algorithm, the particle population composition ring structure is connected with perfect interconnection structure. At the same time in order to prevent the algorithm into local extremum, join in adaptive algorithm escape function, selection of poor fitness particles were redistributed, and into high quality particles in the ring as the object of learning. Generate enough points to re search, to obtain the global optimal value. Through 6 standard test function optimization, compared with other algorithms, the RSEPSO can obviously improve the performance of particle swarm optimization algorithm.

     

/

返回文章
返回