协同免疫克隆算法及其在零等待flowshop调度中的应用

The co-evolutionary immune clone algorithm and its application in zero-wait flowshop scheduling

针对零等待flowshop调度问题,提出一种叫做协同免疫克隆算法(CICA)的新的有效的算法,该算法将克隆选择机制和免疫系统原理结合起来,在局部操作中引入激励度函数和新的亲和突变操作,使得抗体应答抗原的综合能力不仅和其亲和力有关,而且与其浓度也有关系,从而能增强抗体的多样性,防止早熟。在全局操作中加入了协同进化思想,在进化过程中进行精英迁移,以加快算法的收敛速度,最终达到优化的目的。仿真结果表明,该算法的收敛速度明显优于免疫算法和免疫克隆算法,验证了该算法的有效性和优越性。最后仿真讨论了激励度系数和反馈系数的选取对算法的影响。

For solving the zero-wait flowshop problem, the paper proposes an effective approach called co-evolutionary immune clone algorithm （CICA）. The algorithm combines the colonial selection mechanism with the principle of immune system, and adds in a new operational of affinity mutation and a new function of activity in local operation, to make the capacity of antibodies to antigens not only relate with their affinity but also concem with their concentration, consequently improving the diversity of the antibodies and avoiding their prematurity. In collective operation, a thinking of co-evolutionary and the elite migration process in evolution are used to accelerate the convergence and achieve the purpose of optimization. The simulation result demonstrates the search precision of the CICA is more effective and highly advantageous than that of the immune algorithm and the immune colonial algorithm, thus verifies its validation and excellence. Finally, the paper discusses the influence of the incentive coefficient and the feedback coefficient on the algorithm＇s performance by simulation.