Abstract: There is a great deal of knowledge rules in a complex knowledge base and it is difficult to grasp the organizations and the logical relationships of the rules. To overcome such difficulty, a Hierarchical Fuzzy Petri Nets (HFPN) was proposed on the basis of Fuzzy Petri Nets (FPN). The formal definitions, the knowledge representation method and reasoning of HFPN were also developed. By application of place abstraction and transition abstraction HFPN is able to realize hierarchical knowledge representation and reasoning. Using the concept of hierarchy, the complexity of constructing knowledge bases can be reduced by modularization method in the design phase, hence facilitates the knowledge base management and optimization.

Key words: computer engineering, knowledge base, fuzzy logic, hierarchical fuzzy P etri nets