Abstract: Participation in organizational random behavior is a key factor in evaluating effective implementation of business collaboration. Combining process algebra and Markov chain, this paper proposes a method for analyzing random behavior of collaborative business processes. Firstly, a finite state automaton is used to model the business processes of each participating organization, and the collaborative service process is defined by introducing asynchronous message communication relation. Secondly, it is proposed a method of transforming the participating business process into Communication Sequential Process（CSP）, and then the CSP process corresponding to each participating organization business process is concurrently combined to obtain the corresponding CSP process corresponding to the business process, and the state migration model of the collaborative business process is constructed according to the CSP operation semantics. Finally, the concept of regularization of the state transition model is introduced, it is theoretically proved that the regularized state transition model corresponds to a homogeneous Markov chain, and then the stability probability of each state in the state transition system is obtained according to the equilibrium equation, based on this, the stochastic behavior analysis of participating organizations is realized. The feasibility and validity of the method are illustrated by analyzing the process of purchasing and stochastic behavior of electrical equipment.

Key words: collaborative business process, asynchronous message communication, random behavior, Communication Sequential Process（CSP）, Markov chain

摘要： 参与组织随机行为是评价业务协同有效实施的一个关键因素。结合进程代数和马尔科夫链，提出了一种协同业务过程的随机行为分析方法。首先，使用有限状态自动机建模每个参与组织的业务过程，通过引入异步消息通信关系定义协同业务过程。其次，提出将参与组织业务过程转换成通信顺序进程（Communication Sequential Process，CSP）方法，进而将每个参与组织业务过程对应CSP进程并发组合得到协同业务过程对应的CSP进程，并根据CSP操作语义构建协同业务过程的状态迁移模型。最后，引入状态迁移模型正则化概念，从理论上证明正则化的状态迁移模型与一个齐次马尔科夫链相对应，进而根据平衡方程求得状态迁移系统中每个状态的稳定概率，以此为基础实现参与组织随机行为分析。通过对电设备采购过程建模与随机行为分析阐述该方法的可行性和有效性。

关键词: 协同业务过程, 异步消息通信, 随机行为, 通信顺序进程（CSP）, 马尔科夫链