Neural simulation of Petri nets

Abstract Petri nets and neural networks share a number of analogies. Investigations of their relationships can be sorted into two categories: (a) the modeling of neural activities with Petri nets, and (b) the neural simulation of Petri nets. The work presented in this paper belongs to the second category. Unlike divide-and-conquer approaches, the proposed method settles the extraneous skeleton of simulators. Inherent distinctions of Petri nets are characterized by the individual constituents of simulators. The constructed simulators thus reveal a consistently uniform structure on a macroscopic level. Compared with those generated by the divide-and-conquer approaches, ours look much portable and are empirically economic. Furthermore, in a fully parallel machine with enough nodes the overall time complexity of the neural simulator will be constant.