In a previous paper a method for knowledge representation and acquisition in sequential control systems was proposed. The objective of that paper was to simplify the acquisition of knowledge and the transition from the specification of sequential control systems design to the Petri net modeling. The method utilized a Petri-net based representation of the concepts for knowledge acquisition in sequential control systems by learning. During the utilization of the proposed method, a single concept could be acquired from many different cases resulting in different representations for the same concept. These representations may include some common parts as well as distinct parts. The authors propose a method to merge these representations into one unique concept, according to the particular features of sequential control systems.
展开▼