带杂交算子的自适应混合花粉授粉算法

摘要

针对基本花授粉算法存在的易陷入局部最优、求解精度不高等问题,提出一种带杂交算子的自适应混合花授粉算法,分别针对异花授粉过程和自花授粉过程设计自适应权重改进和引入杂交算子.并用六个标准函数测试算法的性能,结果表明MFPA寻优性能优于基本花授粉算法和基于模拟退火的花授粉算法.%To solve the problems of basic flower pollination algorithm easily falling into local optimum and the low computa?tional accuracy,presenting adaptive flower pollination algorithm based on crossover operator,which means designing adaptive weight and crossover operator for global pollination and local pollination. Six standard functions are tested,and the results show that MFPA is better than basic flower pollination algorithm and SFFPA.