基于球面杂交的自适应动态克隆选择算法

Adaptive Dynamic Clone Selection Algorithm Based on Sphere Crossover

在克隆选择算法搜索函数最优解问题的研究中,针对传统自适应动态克隆选择算法收敛速度慢、精度低以及种群多样性低的缺点,提出了一个基于球面杂交的自适应动态克隆选择算法。新算法采用浮点数编码方式,在每次迭代过程中,首先根据抗体的亲和度动态计算出每个抗体的变异概率,然后根据亲和度大小将抗体种群动态分为记忆单元和一般抗体单元,并采用球面杂交方式对种群进行调整,提高了算法的收敛速度和求解精度。实例验证了所提算法的有效性和可行性。

As the traditional adaptive dynamic clone selection algorithm has some shortcomings,such as slow convergence,low accuracy and low population diversity,a new clone selection algorithm for searching optimal solution of function is studied and an adaptive dynamic clone selection algorithm is presented to improve it,which is based on sphere crossover.Firstly,using floating-point encoding,the mutation probability of each antibody is dynamically calculated in each iteration process,according to antibody affinity.And then,according to the size of antibody affinity,antibody populations are dynamically divided into memory antibody units and general antibody units.Subsequently,antibody populations are adjusted by sphere crossover so that algorithm convergence speed and solution accuracy are improved.The effectiveness and the feasibility of the proposed algorithm are verified by examples.