Fuzzy Petri nets (FPNs) are a promising modeling tool for knowledge representation and reasoning. As a new type of FPNs, bipolar fuzzy Petri nets (BFPNs) are developed in this article to overcome the shortcomings and improve the performance of traditional FPNs. In order to depict expert knowledge more accurately, the BFPN model adopts bipolar fuzzy sets (BFSs), which are characterized by the satisfaction degree to property and the satisfaction degree to its counter property, to represent knowledge parameters. Because of the increasing scale of expert systems, a concurrent hierarchical reasoning algorithm is introduced to simplify the structure of BFPNs and reduce the computation complexity of knowledge reasoning algorithm. In addition, a large group expert weighting method is proposed for knowledge acquisition by taking experts’ non-cooperative behaviors into account. A realistic case of risk index evaluation system is presented to show the effectiveness and practicality of the proposed BFPNs. The result shows that the new BFPN model is feasible and efficient for knowledge representation and acquisition.

中文翻译：

---

考虑非合作行为的双极模糊Petri网用于知识表示和获取

模糊Petri网（FPN）是用于知识表示和推理的有前途的建模工具。本文开发了一种双极性模糊Petri网（BFPN）作为新型的FPN，以克服传统FPN的缺点，提高其性能。为了更准确地描述专家知识，BFPN模型采用双极模糊集（BFS），以对属性的满意程度和对其相对属性的满意程度为特征，来表示知识参数。由于专家系统规模的不断扩大，引入并发分层推理算法以简化BFPNs的结构并降低知识推理算法的计算复杂度。此外，提出了一种考虑专家的非合作行为的知识获取大群体专家方法。提出了一个现实的风险指标评估系统案例，以证明所提出的BFPN的有效性和实用性。结果表明，新的BFPN模型对于知识的表示和获取是可行和有效的。