基于直觉模糊Petri网的混合推理方法

发布时间：2018-05-23 20:17

本文选题：直觉模糊Petri网 + 正向推理　；参考：《电子学报》2017年08期

【摘要】：针对现有的基于模糊Petri网(Fuzzy Petri Nets,FPN)和直觉模糊Petri网(Intuitionistic Fuzzy Petri Nets,IFPN)的推理方法在求解只涉及知识库中部分规则的问题时存在推理过程复杂、效率不高,而且不能对问题产生的原因进行分析等缺陷,提出一种基于IFPN的混合推理方法.该方法将反向推理与正向推理相结合,首先把所要求解的问题转化为目标库所,并引入关联库所、关联变迁和子模型等概念;其次运用反向推理寻找目标库所的关联库所和变迁并构建推理子模型,从而获取问题产生的潜在原因并简化推理模型;最后以子模型作为推理模型,运用正向推理求解目标库所的token值,解决了直接运用原模型进行推理时过程复杂且效率不高的问题.与此同时,通过在模型中引入阈值以及"路径"和"有效路径"等定义,排除无效关联库所,从而找出了问题产生的真正原因.实例验证表明该方法可行且有效,与现有方法的对比分析表明该方法克服了现有方法的缺陷.
[Abstract]:The existing reasoning methods based on fuzzy Petri nets and intuitionistic Fuzzy Petri networks have complex reasoning process and low efficiency in solving problems involving only some rules in the knowledge base. Furthermore, a hybrid reasoning method based on IFPN is proposed, which can not analyze the causes of the problem. This method combines reverse reasoning with forward reasoning. Firstly, the problem required to be solved is transformed into a target library, and the concepts of association library, association transition and sub-model are introduced. Secondly, reverse reasoning is used to find the relevance database and the transition of the target library and construct the inference submodel, so as to obtain the potential causes of the problem and simplify the reasoning model. Finally, the submodel is used as the reasoning model. By using forward reasoning to solve the token value of the object library, the complex and inefficient process of direct reasoning with the original model is solved. At the same time, by introducing the threshold and the definitions of "path" and "effective path" in the model, the invalid association library is excluded, and the real cause of the problem is found out. The example shows that the method is feasible and effective, and the comparison with the existing methods shows that the method overcomes the shortcomings of the existing methods.
【作者单位】：空军工程大学防空反导学院;武警工程大学;
【基金】：国家自然科学基金(No.61272011) 国家自然科学青年基金(No.61309022)
【分类号】：TP181;TP301.1

【相似文献】