一种以超球体作为抗体识别区域的新的实值阴性选择算法

摘要

In the view of geometry model, the function of artificial immune is to divide the space into areas.Hypersphere is regarded as an antibody recognition zone of artificial immune system. The shortage of the traditional negative selection algorithm is discussed, and an improved algorithm is presented. Simulation and experimental results indicate that the proposed method is feasible.%从几何模型出发,免疫问题实际上是一个空间划分的问题.将超球体作为免疫系统抗体识别区域,针对经典阴性选择算法的不足,提出了改善的方法,从根本上减少了误报和漏报率,降低了检测器的冗余可能性.仿真和实验实例表明,该算法适合于电路板的故障诊断,有较高的故障诊断率.