基于Fuzzy Petri网专家系统的研究

提出采用ＦｕｚｚｙＰｅｔｒｉ网作为知识表示和推理机制的模型，发挥其较好的模型描述和能处理并行推理的特点，基于ＦｕｚｚｙＰｅｔｒｉ网构造了一个耦合结构的专家数据库系统（ＥＤＢＳ），提高了推理的效率及其准确性，同时具有并发控制，数据组织，共享，存取和查询等数据库管理系统的功能，并使用投资决策实例对模型进行了解释和验证。

On Expert Database System Based on Fuzzy Petri Net

Investment companies in China now need decision support systems (DSS). We developed an Expert Database System (EDBS) as a part of one particular type of DSS. We use a loosely coupled structure of Expert System(ES) and Data Base Management System (DBMS) to construct an EDBS based on fuzzy petri net model. The EDBS consists of ES, DBMS, humancomputer interface and communication interface. Through an example of using our EDBS as part of a DSS in an actual investment analysis and decision, we explain and test our EDBS. In Expert System of our EDBS, we make use of fuzzy petri net model for the knowledge representation. The fuzzy, concurrent, and asynchronous reasoning is implemented, and the reasoning efficiency and accuracy are improved through the computation of confidence and through the algorithms for fuzzy reasoning based on fuzzy petri net model. In DBMS of our EDBS, all sorts of management functions and graphs have been implemented. The management functions include concurrency control, data organization, storage, fault recovery, and query.