Optimizing Fuzzy Petri Nets by Using an Improved Ant Colony Algorithm

It is very important for constructing a FPN (fuzzy petri net) to accurately find out all parameters of fuzzy production rules. In this paper, Maximum-Minimum Ant System (MMAS) of ant colony algorithm (ACA) is originally introduced into the process of exploring the optimal parameters of a modified FPN.The optimization algoritnm is based on the techniques of multithreading. Realization of the algorithm do not depend on experiential data and requirements for the initial input of the FPN are not stringent. Simulation experiment shows that the parameters trained by the above MMAS multithreading algorithm are highly accurate and the FPN model constructed by these parameters possesses strong generalizing capability and self-adjusting purpose.