【摘要】 提出一种基于因果关系的思想求解所有极小诊断的新方法。此方法与Reiter的模型诊断方法有着本质的不同,不用求解冲突集及冲突集的碰集,一次直接求出所有的极小诊断,避免了由于调用ATMS时的组合爆炸引起的NP-完全问题。并根据因果关系的思想提出了极小诊断组成原理和极小诊断元件个数原理,可直接求出所有的极小诊断,避免了非极小诊断的产生,提高了诊断效率,并且不会丢失正确的解。该算法程序容易编制,且效率较好,可以满足复杂被诊断对象的实时性要求。更多还原

【Abstract】 A novel method of model-based diagnosis by using causal relations was proposed for computation of all minimal diagnoses.This approach is substantially different from the classical model-based diagnosis methods.Unlike the classical methods,it can directly obtain all the minimal diagnoses without computing all the conflict sets and therefore the hitting sets of the collection of the corresponding conflict sets.The NP-complete problems induced by combinatorial explosion due to calling ATMS can be avoided.Based on the principle of causality the composing mechanism of minimal diagnosis and the theory of the number of elements in a minimal diagnosis were put forward.As a result,the non-minimal diagnoses can never be produced,and the true solution can not be missed.The algorithm is easy to program,and the proposed method can significantly improve the diagnosis efficiency to satisfy real-time requirement even for a complex system.更多还原