Synthesis of intelligent switching systems using neural Petri nets

An intelligent switching system is a structure that is able to evaluate criteria of the system and dynamically make a selection from among a discrete set of options. The goal of our research is to develop a system for distributed control, where the control activity is governed by a set of devices that cooperatively learn and make decisions. The test bed for this system is a computer-controlled model railroad. In this test paradigm, each track switch is an intelligent switching device. A neural Petri net is used to model the switch. This paper presents the design and development of the switch, and a modeling of the railroad.