从生物体新陈代谢的生理机能出发,建立了基于酶催化的人工代谢算法模型.通过对代谢算子功能的分析提出了模式定理的概念,并证明了算法的收敛性.利用酶与底物相匹配的原理,通过代谢操作找出酶与底物的最优匹配,从而找出对应故障现象的原因.在寻优过程中,不断地通过凋亡算子将不可能成立的原因筛除,降低了故障诊断的成本.实例分析表明,该算法能有效地进行故障定位,尽可能地缩小了不必要的搜索空间.
Artificial metabolic algorithm (AMA) model is built based on enzyme catalysis from the idea of biological metabolism. The schema theorem is proposed by discussing metabolic operators function, and convergence of the algorithm is proved. Taking advantage of matching principle between substrate and enzyme, the optimal matching relation between specified enzyme and specified substrate is found by metabolic operation. The optimized result corresponds to reason for fault. In optimization process, impossible reasons are eliminated continuously by apoptosis operator. The cost for fault diagnosis is decreased. An example analysis shows that fault can be located by the algorithm efficiently and the unnecessary search space can be omitted as much as possible.