基于PMC模型的高效人工免疫诊断算法

A PMC MODEL BASED EFFICIENT ARTIFICIAL IMMUNE DIAGNOSIS ALGORITHM

诊断算法是系统级故障诊断的一个重要研究课题。文献[10]提出PMC模型下的人工免疫诊断算法,但是没有充分利用PMC模型的特点,并在计算亲和度的时候给出一个假设。对此首先利用PMC模型的特点优化初始种群的质量,其次定义优化后的亲和度函数,最后提出新的算法流程,并证明算法的正确性和收敛性。模拟实验显示在CPU时间和迭代次数上都优于原算法。

Diagnosis algorithm is an important research subject in system-level diagnosis.Reference document proposed an artificial immune diagnosis algorithm under PMC model,but failed to take full advantage of PMC model,whereas made a hypothesis in the computing affinity.Therefore the paper firstly optimizes the quality of initial population by utilizing PMC model characteristics;then defines the optimized affinity function;at last proposes a new algorithm process.The algorithm is validated for correctness and convergence.Simulation experiment shows it is superior to the original algorithm at both CPU time and iteration times.