一种精英反向学习的花授粉算法

A flower pollination algorithm based on elite opposition-based learning

针对花授粉算法(Flower Pollination Algorithm,FPA)在处理高维优化问题时收敛精度低、速度慢等缺陷,提出一种基于精英反向学习的花授粉算法(Elite Opposition-Based Learning Flower Pollination Algorithm,EOBLFPA).改进的算法使用全局最优花朵的反向解更新花朵个体的位置,引导花朵个体向最优解靠近.在10个经典测试函数上做仿真实验,结果表明,EOBLFPA算法在收敛精度和收敛速度上有显著提升,寻优性能优于FPA算法和相关对比算法,并验证了其改进策略的有效性.

A flower pollination algorithm(FPA)based on elite opposition-based learning is presented to overcome the problems of low convergence precision and low convergence speed in dealing with high-dimensional optimization problems.The opposite solution of the best flower was used to update a flower′s position.This strategy enables the flower individuals to be guided near the optimal solution.Experiments were conducted on ten benchmark test function.The results demonstrate a better performance of the EOBLFPA than basic flower pollination algorithm and the improved algorithm.The convergence precision and convergence speed of the al-gorithm were improved.Therefore,the modified algorithm is proved to be effective.