Desenvolvimento de uma metodologia em redes de Petri a objetos interpretada

Abstract

In the 1960s, the mathematician and computer scientist C. A. Petri developed an important tool that allows events previously described only by temporal relations to be described by causal relations. Petri Nets (PN) formalism allows the representation of discrete event systems (DES), which are described by the sequence of events that can occur and the sequence of states that can be visited. For more than 50 years, inspired by Petri, scientists have created other net definitions that can aggregate more information and thus represent more complex systems. Such nets are able to represent more complex systems, which previously could not be described because of the explosion of states of finite state automata. Two of these are particularly important: (i) Signal Interpreted Petri Net (SIPN), capable of receiving input signals, processing them, and then providing output signals for a Programmable Logic Controller (PLC) to execute a method; (ii) Object Petri Net (OPN), capable of treating the tokens as individual objects that carry information, with different attributes and methods. These net definitions are not always able to describe a proposed problem when used separately. This work is based on the concepts of places, transitions, arcs, weight and marking function of a PN (including graphic notation), object orientation (high level) and signal interpretation (low level). The objective of this work is to formalize a new definition of Petri net that incorporates the above concepts: the Signal Interpreted Object Petri Net (SIOPN). The developed notation is applied in a case study for polishing and separating recyclable materials (metal, wood and black plastic), since recycling is a prominent topic today and may have a practical industrial application. The case study begins with all blocks in their raw state in an initial buffer. The separation is done by reading a logical combination of low level signals from inductive, optical and capacitive sensors. Once the material is identified, such information is incorporated into the object and appropriate actions are taken when and if necessary. At the end of the process, all materials will be polished and separated into their respective buffers. The implementation of SIOPN shown in this case study serves as an example for new implementations to be carried out in future works, respecting the particularities of each system.