/ informatics / theory /

[edit]

Definition

A petri net N=(P,T,E)N = (P,T,E) is a graph to model distributed systems. It consists of places PP and transitions TT, which are connected by directed edges EE.

The places are marked with tokens. Tokens are transported by the edges. A transition consumes and produces tokens. The state of the net is defined by the numbers of tokens at each place.

The initial marking M0 defines the number of tokens at each place.

Types

Execution Rules

The rules for a boolean, non-weighted petri net:

In a weighted Petri Net, the transition consumes and produces tokens according to the weights of the edges.

Petri Net Transition
Petri Net Transition

Properties

If a petri net is live with initial marking M0, then it is deadlock-free

Context

Petri nets are more general than state machines because they can model concurrency. A state machine is a special petri net where each transition is only connected to one input and one output place (no concurrency).

Petri Net Classes
Petri Net Classes