Enterprise-level business component identification in business architecture integration

The component-based business architecture integration of military information systems is a popu- lar research topic in the field of military operational research. Identifying enterprise-level business components is an important issue in business architecture integration. Currently used methodologies for business component identification tend to focus on software-level business components, and ignore such enterprise concerns in business architectures as organizations and resources. Moreover, approaches to enterprise-level business component identi- fication have proven laborious. In this study, we propose a novel approach to enterprise-level business component identification by considering overall cohesion, coupling, granularity, maintainability, and reusability. We first define and formulate enterprise-level business components based on the component business model and the Department of Defense Architecture Framework （DoDAF） models. To quantify the indices of business components, we formulate a create, read, update, and delete （CRUD） matrix and use six metrics as criteria. We then formulate business com- ponent identification as a multi：objective optimization problem and solve it by a novel meta-heuristic optimization algorithm called the ＇simulated annealing hybrid genetic algorithm （SHGA）＇. Case studies showed that our approach is more practical and efficient for enterprise-level business component identification than prevalent approaches.

基于组件的军事信息系统业务架构集成是军事领域中一个重要研究内容,而识别企业级业务组件是业务架构集成中一个关键问题。当前业务组件识别的方法多是关注于软件层面业务组件,忽略了诸如组织、资源等企业级因素;而目前企业级业务组件识别方法被证明非常低效。因此本文提出一种企业级业务组件识别的新方法,该方法全面考虑了业务组件的内聚度、耦合度、粒度、可维护性、可复用性五个设计原则。首先基于业务组件模型和DoD AF(Department of Defense Architecture Framework)模型对业务组件进行了定义和形式化描述,为了对业务组件进行定量化分析,将业务模型转为一个CRUD(create,read,update,and delete)矩阵并提出了6类指标;然后将业务组件识别问题转化为一个多目标优化问题,并采用了模拟退火遗传算法(simulated annealing hybrid genetic algorithm,SHGA)进行求解。最后通过案例分析验证了本文方法较先前的方法对企业级业务组件识别具有更好的适用性和高效性。