Petri võrkude teisendamine BPMN mudeliteks
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Antud magistritöö käsitleb Petri võrkude teisendamist samaväärseteksBPMN mudeliteks. Täpsemalt öeldes keskendub see Petri võrkude alamklassilenimega töövoo võrgud. Antud lõputöös implementeeriti teisendaja, kasutadesselleks mitmeid tehnikaid nagu näiteks Petri võrkude lahti pakkimineja modulaarsed dekompositsiooni puud. Sellest tulenevalt pakub antud magistritöö välja täieliku teisendusalgoritmi, mis suudab käsitleda sümmeetrilisi segadusi Petri võrkudes. See on antud valdkonnas üks esimesi teisendamise meetodeid, mis katab ka seda klassi. Hetkel oleme teadlikud ainult ühest teosest, mis illustreerib mõlemasuunalist teisendamist töövoo võrkude ja graafide vahel. Lisaks, esitleme me käitumissõltuvuste maatriksi arvutamise meetodi.Käsitleme ka erijuhtumeid, kus peame BPMN-is lisama tau sündmuse, ettegemist oleks samakujulise mudeliga.
This thesis addresses the problem of translating a Petri net into an equivalentBPMN process model. This is fundamental problem with implications on theunderstanding of the semantics of the notation and that has potential applications in areas such process model discovery from event logs and structuring of process models. In previous work, it has been shown that the well-known family of free-choice Petri nets can be bidirectionally mapped into the subset of BPMN process models constructed solely with tasks and exclusive/parallel getaways. In contrast, this work searches at lifting the restriction to a larger family of Petri nets by proposing a translation that covers also the case of nets with symmetric confusion. The approach has been implemented in a prototype which has allowed us to conduct a preliminary performance study.
This thesis addresses the problem of translating a Petri net into an equivalentBPMN process model. This is fundamental problem with implications on theunderstanding of the semantics of the notation and that has potential applications in areas such process model discovery from event logs and structuring of process models. In previous work, it has been shown that the well-known family of free-choice Petri nets can be bidirectionally mapped into the subset of BPMN process models constructed solely with tasks and exclusive/parallel getaways. In contrast, this work searches at lifting the restriction to a larger family of Petri nets by proposing a translation that covers also the case of nets with symmetric confusion. The approach has been implemented in a prototype which has allowed us to conduct a preliminary performance study.