摘要
基于生物免疫原理提出了一种故障诊断方法,将故障作为抗原,首先应用反面选择算法产生初始抗体,然后基于生物免疫系统的克隆、变异、克隆选择、自我调节和免疫记忆等机理对初始抗体进行进化,产生能够表示和识别抗原即故障的记忆抗体,并以抽油机井为对象进行了仿真研究,结果表明,该方法能够取得很好的诊断效果。
A fault diagnosis method based on the biology immune principle is developed, in which fault is regarded as antigen, first initial antibodis are generated using negative selection algorithms, then based on clone, mutation, clone selection, self adjustment and immune memory of the biology immune system, the initial antibodies is evolved to generate memory antibodies which can express and identify the antigen or fault. The simulations are done through the experiments of the pump-jack and the results show that good fault diagnosis effect can be achieved.
出处
《计算机测量与控制》
CSCD
2006年第3期287-288,298,共3页
Computer Measurement &Control
基金
黑龙江省自然科学基金资助项目(TF2005-26)
关键词
故障诊断
反面选择
克隆选择
抗体
抗原
fault diagnosis
negative selection
clone selection
antibody
antigen