CAJ | 학술논문

提出了基于混合数据类型的免疫聚类算法，将零件各设计参数作为抗体的基因位，通过聚类分析中设定的阈值将存在设计缺陷的抗体直接剔除进死亡抗体库，从而实现了对抗体设计缺陷的辨识。将最优设计中经交叉变异产生的抗体的双亲作为存在设计缺陷的抗体并入死亡抗体库。
제출료기우혼합수거류형적면역취류산법，장령건각설계삼수작위항체적기인위，통과취류분석중설정적역치장존재설계결함적항체직접척제진사망항체고，종이실현료대항체설계결함적변식。장최우설계중경교차변이산생적항체적쌍친작위존재설계결함적항체병입사망항체고。
This paper put forward immune clustering algorithm based on mixed data types, parts of the design parameters as antibody gene, through cluster analysis in setting the threshold will exist design flaw antibody eliminated directly into the death of antibody library, in order to achieve the design defects of antibody identification. This paper design the optimal de-sign of the cross variation of antibodies generated by the parents as a design defect of the antibody incorporated into the death of the antibody library.