CAJ | 학술논문

针对基本蝙蝠算法在寻优后期存在寻优精度低、早熟及易陷入局部最优等缺陷，提出一种融合入侵杂草生长繁殖局部搜索的蝙蝠算法。该算法将入侵杂草算法的杂草生长繁殖、空间扩散和竞争机制融合到蝙蝠算法中，采用动态调整杂草空间扩散算子的标准差，使算法既增加了全局搜索能力，同时也提高了局部搜索能力。通过对8个标准测试函数的仿真测试，结果表明该算法具有较优的全局搜索和局部搜索能力，在寻优精度和收敛性方面都比基本蝙蝠算法有较大的提高。
침대기본편복산법재심우후기존재심우정도저、조숙급역함입국부최우등결함，제출일충융합입침잡초생장번식국부수색적편복산법。해산법장입침잡초산법적잡초생장번식、공간확산화경쟁궤제융합도편복산법중，채용동태조정잡초공간확산산자적표준차，사산법기증가료전국수색능력，동시야제고료국부수색능력。통과대8개표준측시함수적방진측시，결과표명해산법구유교우적전국수색화국부수색능력，재심우정도화수렴성방면도비기본편복산법유교대적제고。
In its later period of optimisation,original bat algorithm has the defects of low accuracy,prematurity and easy to fall into local optimum,etc.To overcome them,we present such a bat algorithm which fuses invasive weed growth and reproduction.It fuses the weed growth and reproduction,spatial diffusion and competition mechanism of invasive weeds algorithm into bat algorithm and dynamically adjusts the standard deviation of weeds spatial diffusion operator,thus the algorithm increases the global and local search abilities both.Through the simulation testing on 8 standard test functions,the results show that this algorithm has better global and local search abilities,in the aspects of optimisation accuracy and convergence it gains greater improvement than the basic bat algorithm.