مدل سازی پتری رنگی زماندار شبکه ها برای جریان کاری پویا در فرایند توسعه محصول
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|21807||2008||17 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Computers in Industry, Volume 59, Issues 2–3, March 2008, Pages 193–209
Product development process (PDP) is one of the most important business processes for enterprises but it has difficulty in workflow management because of the uncertain and dynamic characteristics. Thus, even though there have been many workflow modeling and management methods, they have limitations to deal with the special characteristics of PDP. We define PWF-nets based on timed colored Petri nets for the workflow management of PDP and propose a method of organizing PWF-nets that are composed of workflow patterns. Considering the characteristics of PDP, PWF-nets are somewhat complex unlike existing workflow modeling because they manage necessary information related to dynamic workflow execution. Thus, we define transition behaviors and function primitives to support the construction of workflow patterns for PWF-nets. These workflow patterns can be microscopically handled and can be adapted for the dynamic and uncertain environment related to data and time. Next, we discuss the soundness and properties of the PWF-net and introduce an example of implementation. Additionally, we can check errors or abnormal executions of workflow and evaluate the lead time of PDP through our program for the PWF-net
The companies have numerous business processes and a key success factor in competition is the effective management of core business processes. In this context workflow technology has been introduced for systematic business process management that supports a consistent information flow between agents and the adequate sharing the data and information. There have been many works related to the workflow applications. Most of works focused on general business processes which are relatively simple and repetitive so many workflow management systems (WFMS) adequately support when the business process to be executed is well-structured. In these cases, we can predefine all reachable paths in advance. However, PDP is hard to be planned completely a priori because the detail route of workflow can be generally determined at run-time. There are several studies focused on the PDP with the special characteristics such as uncertain, evolutionary, iterative, and coupled. Petri net allows a smooth transition from the conceptual level of process model to an implementation level  and it is a state-based description so it has been broadly used for workflow modeling. van der Aalst  explained the reason of using a state-based description that can keep distinct between the enabling of an activity and the execution of an activity. There are various types of Petri net related to workflow modeling. Basic Petri nets describing a real system tends to be complex and extremely large so that extended Petri nets with time or color has been used to model a system . Extended Petri nets include several high-level Petri nets such as a timed Petri net, a colored Petri net, and a hierarchical Petri net. High-level Petri nets are suitable to workflow modeling of PDP because they can manage various attributes like due date and whether some predecessor activities are executed or not. A timed Petri net generally associates transitions with deterministic or stochastic firing times so that we can analyze the temporal behavior of PDP using them. In a colored Petri net, attributes are associated with tokens and each activity is assigned to the tokens of own types. Especially, within a workflow modeling, the color or the value of a token contains data about the case such as routing parameter, due date, responsible or preferred resource . These data are necessary to PDP management and attribute of them can be updated according to the progress of workflow. Lastly, a hierarchical Petri net allows for hierarchical modeling so it is possible to decompose a complex system into simpler subsystems. There have been many Petri net-based workflow modeling methods but they mainly focused on the quite simple and repetitive workflow. Therefore they have difficulty in supporting the characteristics of PDP and require the additional consideration. In this paper, we define a PWF-net that is based on a timed colored Petri net for the workflow management of PDP and additionally propose a method of composing a PWF-net by workflow patterns. First of all, we introduce Petri net-based workflow patterns for a PWF-net. These patterns represent the typical execution types of PDP and can support a less rigid workflow specification for PDP which contains the workflow of fixed route and ad hoc route as well. The fixed part of workflow is specified at build-time but the ad hoc part of workflow cannot be wholly specified at build-time, and it can be completely determined during run-time. Workflow patterns are composed of six types such as Sequential, Overlapped, Iterative, Coupled, Split, and Join pattern. Considering the characteristics of PDP, PWF-nets are somewhat complex unlike existing workflow modeling because has to manage necessary information that is related to the execution of dynamic PDP workflow. Thus, we define four transition behaviors and eight function primitives to support the construction of workflow patterns for PWF-nets. Transition behaviors are related to the cases that the color of token are changed and function primitives represent the types that are several typical structure included by workflow patterns. Workflow patterns formed by function primitives and transition behaviors are microscopically handled and can be adapted for the dynamic and uncertain environment related to data and time. Next, we discuss the soundness and the properties of PWF-net and introduce an example of implementation. Additionally, we can check errors or abnormal executions of workflow and evaluate the lead time of PDP through PWF-net executor. The rest of this paper is organized as follows. In the following chapter, related works are introduced. In Section 3, modeling patterns for workflow of PDP are introduced. Petri nets for workflow model are introduced and Petri net for PDP workflow (PWF-net) are suggested in Section 4. Analysis of PWF-net and related program and implementation are presented in Section 5. Finally, conclusion is given in Section 6.
نتیجه گیری انگلیسی
Petri nets have been broadly used for workflow modeling and high-level Petri nets are suitable for complex PDP workflow. A timed Petri nets can analyze the temporal behavior of system and a colored Petri nets can dealt with objects with various attributes. In this context, we defined PWF-nets that are based on a timed colored Petri nets for the workflow management of PDP. PWF-nets that apply relatively many colors are very complex so it is difficult to compose the proper PWF-nets. To easily get PWF-nets, we proposed a method of composing PWF-nets with workflow patterns in detailed. These patterns mean the typical execution types of PDP. Especially, we introduced the process template for the dynamic and uncertain characteristic of workflow because PDP contains a lot of ad hoc route. The ad hoc part of workflow cannot be whole specified at build-time and it is completely determined during run-time. The workflow patterns for PWF-nets are somewhat complex because each pattern has to manage necessary information that is related to the execution of dynamic PDP workflow. In behalf of this adversity, we designed transition behaviors and function primitives to support the construction of workflow patterns. Consequently, workflow patterns in this paper can be microscopically handled and adapted the dynamic and uncertain environment. Last, we discussed the soundness and introduced an example of implementation. Basically, we designed the semi-structured frame that can accommodate the appropriate variation because the implementation of system for dynamic workflow is restricted. Practically, this frame using process templates deals with the predictable change of workflow and it can support to systematically analyze the various changes of workflow.