蝙蝠算法的一种改进方法

An improved method for bat algorithm

针对蝙蝠算法在进行局部搜索时,易使算法陷入局部极值的束缚,导致算法收敛精度不高的缺陷,提出了使用t-分布对局部搜索时的最优解进行变异操作.为最优解各维度增加t分布型随机扰动项,选取7个经典测试函数做仿真实验.实验结果表明:改进的蝙蝠算法在收敛精度和速度上有显著提升,说明通过对最优解实施t-分布扰动能够使算法摆脱局部极值的束缚,显著提高收敛精度.

When the bat algorithm performs local search, random numbers are added to the optimal solution of each dimension. This mechanism makes the bat algorithm fall into local extremum, which leads to the low precision of the algorithm. In view of the shortcomings of the bat algorithm. This paper proposes a modified algorithm, which employed a student＇s t-distribution mutation operator to disturb the optimal solution of each dimension. The experimental results of seven function show that the modified algorithm improves the convergence precision and speed. Therefore, the modified algorithm which t-distribution mutation operator is added can improve the abilities of seeking the global excellent result and evolution speed.