Extended Fuzzy Petri Net for multi-stage fuzzy logic inference

This paper presents a new extended Fuzzy Petri Net (EFPN) model to demonstrate the dynamic behavior of fuzzy inference systems during their executing process. The proposed EFPN model is based on our previously proposed fuzzy reasoning algorithms and implemented in GUI environment to support visual views of the dynamic behavior. The EFPN components and function are mapped from the different type of fuzzy operators of if-parts and then-parts in fuzzy rules. Further, only active fuzzy operators of active fuzzy rules are shown in the GUI windows to present the dynamic behavior. On the other hand, low complexity of the entire proposed EFPN model is available due to the mapping from the proposed fuzzy reasoning algorithms. This proposed method can decrease computation time and provide clear and global views of the dynamic behavior to the designers to reduce their design effort. The EFPN model has been applied to control a truck-and-two-trailer system. The simulation results showed that the proposed method is feasible.