摘要
在业务流程协作过程中,多实例参与者间基于消息进行交互。由于参与者之间的利害关系以及消息传输的不稳定性,存在消息被篡改或延迟到达的情况,从而影响协作的准确性。此外,对内部数据的不当操作也会阻碍协作的有效进行。因此,检测流程内和跨流程的数据流错误至关重要。通常使用Petri网来检测这些数据流错误,然而Petri网缺乏对多实例的定义,若采用增加流程数量来表示增加的实例,则容易出现状态空间爆炸问题。同时Petri网无法表示控制流可达而数据流不可达的状态,无法检测数据缺失问题。为此定义了一种多实例着色Petri网(MCPN),其不但提供了对多实例的形式化描述,而且将控制流和数据流区分表示,基于MCPN能够检测包括数据缺失在内的多种数据流错误。同时也提出了一种基于MCPN的约简算法来缓解状态空间的增长。最后,通过实验证明了提出方法的有效性和优越性。
The participants in Business Process Collaboration(BPC)with multi-instance frequently engage in message-based interactions.Due to the conflicts among participants and the instability of message transmission,there are some situations such as message redundancy or loss,which can impact the accuracy of collaboration.Additionally,improper operations on the internal data hinder the effective collaboration.Therefore,it is essential to detect the data flow errors within and across processes.Petri nets are widely used to detect dataflow errors;however,they lack a formal definition for multi-instance scenarios.Increasing the number of processes to represent multi-instance may lead to state space explosion.Furthermore,Petri nets cannot describe situations where control flow is reachable while data flow remains unreachable,resulting in an inability to detect missing data.Therefore,a Multi-instance Colored Petri Net(MCPN)was introduced,which not only provided a formalized description of multi-instance,but also clearly distinguished between the representation of control and data flow.By utilizing MCPN,it became possible to detect various data flow errors,including missing data.A reduction algorithm based on MCPN was proposed to mitigate the exponential expansion of the state space.Finally,the effectiveness and superiority of the proposed method were demonstrated through some experiments.
作者
贾畅
潘茂林
余阳
JIA Chang;PAN Maolin;YU Yang(School of Computer Science and Engineering,Sun Yat-Sen University,Guangzhou 510006,China;School of Software Engineering,Sun Yat-Sen University,Zhuhai 528406,China)
出处
《计算机集成制造系统》
北大核心
2026年第3期990-998,共9页
Computer Integrated Manufacturing Systems
基金
国家自然科学基金资助项目(61972427)
NSFC-广东联合基金资助项目(U20A6003)
广东省科技计划资助项目(2020A0505100030)。
关键词
业务流程协作
数据流错误
流程内外
多实例参与者
business process collaboration
data flow errors
within and across processes
multi-instance participants
