Reduction rules for reset workflow nets

When a workflow contains a large number of tasks and involves complex control flow dependencies, verification can take too much time or it may even be impossible. Reduction rules can be used to abstract from certain transitions and places in a large net and thus could cut down the size of the net used for verification. Petri nets have been proposed to model and analyse workflows and Petri nets reduction rules have been used for efficient verification of various properties of workflows, such as liveness and boundedness. Reset nets are Petri nets with reset arcs, which can remove tokens from places when a transition fires. The nature of reset arcs closely relates to the cancellation behaviour in workflows. As a result, reset nets have been proposed to formally represent workflows with cancellation behaviour, which is not easily modelled in ordinary Petri nets. Even though reduction rules exist for Petri nets, the nature of reset arcs could invalidate the transformation rules applicable to Petri nets. This motivated us to consider possible reduction rules for reset nets. In this paper, we propose a number of reduction rules for Reset Workflow Nets (RWF-nets) that are soundness preserving. These reduction rules are based on reduction rules available for Petri nets [19] and we present the necessary conditions under which these rules hold in the context of reset nets.