人机合作的免疫算法及其在布局设计中的应用

Human-computer Cooperative Immune Algorithm and Its Application to Layout Design

提出一种人机合作的免疫算法(HCCIA)。它将设计者提供的初始人工个体和混沌初始化生成的初始算法个体合并组成初始群体,然后经自适应交叉、变异和免疫选择等算子操作,进化寻优。同时,还适时和适量地将设计者给出的新人工个体加入到算法群体,替换掉其中较差的个体。如此循环,直至达到结束准则。该算法有利于充分发挥人机各自特长,为更好地解决复杂工程设计问题提供了方便。文中以卫星舱布局设计为背景的算例验证了此算法的可行性和有效性。

This paper proposes a human-computer cooperative immune algorithm （HCCIA）. This algorithm is based on improved immune algorithm and the human intelligence is introduced into it further. Its initial population consists of the initial artificial individuals supplied by human and the initial algorithm individuals generated by chaos initialization. And all the individuals are put into algorithm operation （including adaptive crossover and mutation, and immune selection, etc） to evolve the population to the next generation. Appropriate numbers of new artificial individuals are also introduced and to replace the inferior individuals of the population at an opportune moment during evolution process as well. The procedure of evolutionary computation and interaction are recycled until the end criterion is satisfied. HCCIA benefits giving free rein to the taIents of designers and computers and contributes to solving complex engineering design problems more satisfactorily. An example that originates from the layout design of satellite modules shows that HCCIA is feasible and effective.