【摘要】 故障诊断专家系统的推理策略取决于诊断知识的表达方式。基于模糊Petri网的知识表达是库所、命题、变迁、输入、输出及其可信度、影射等的集合。基于模糊产生式规则的模糊Petri网模型,能较好地表示和处理模糊知识,并能根据变迁的可信度大小,采用深度优先的搜索方法,以反向并行推理来缩小搜索空间,寻找到一条最有可能产生已知故障的路径。为此,以机车柴油机故障信息诊断系统为例,针对其故障信息的不确定性和模糊性,给出了基于Petri网的模糊诊断知识表达方法,讨论了建立在模糊Petri网模型基础上的反向模糊推理算法。更多还原

【Abstract】 The reasoning strategy of fault diagnosis expert system depends on the expression way of diagnosis knowledge. The knowledge representation of fuzzy diagnosis based on petri nets includes knowledge-base station (KBS), proposition, variance, input and output, and reliability and allusiveness. The fuzzy knowledge was preferably expressed and processed with model of fuzzy petri nets baesd on fuzzy rule. According as the believable degree of variance, searching range was reduced with searching method of depth priority and the reverse parallel reasoning, an approach that possibly discover known failure was found. Taking fault diagnosis system for vehicle engine as an example, the expressive method of fuzzy diagnosis knowledge based on petri nets was presented aiming at the uncertainty and fuzziness of the fault messages. The reverse reasoning algorithm baesd on the model of fuzzy petri nets has also been discussed Finally.更多还原