本文選題：業(yè)務架構(gòu)集成 + 業(yè)務組件�。� 參考：《Frontiers of Information Technology & Electronic Engineering》2017年09期

【摘要】：基于組件的軍事信息系統(tǒng)業(yè)務架構(gòu)集成是軍事領(lǐng)域中一個重要研究內(nèi)容,而識別企業(yè)級業(yè)務組件是業(yè)務架構(gòu)集成中一個關(guān)鍵問題。當前業(yè)務組件識別的方法多是關(guān)注于軟件層面業(yè)務組件,忽略了諸如組織、資源等企業(yè)級因素;而目前企業(yè)級業(yè)務組件識別方法被證明非常低效。因此本文提出一種企業(yè)級業(yè)務組件識別的新方法,該方法全面考慮了業(yè)務組件的內(nèi)聚度、耦合度、粒度、可維護性、可復用性五個設計原則。首先基于業(yè)務組件模型和DoD AF(Department of Defense Architecture Framework)模型對業(yè)務組件進行了定義和形式化描述,為了對業(yè)務組件進行定量化分析,將業(yè)務模型轉(zhuǎn)為一個CRUD(create,read,update,and delete)矩陣并提出了6類指標;然后將業(yè)務組件識別問題轉(zhuǎn)化為一個多目標優(yōu)化問題,并采用了模擬退火遺傳算法(simulated annealing hybrid genetic algorithm,SHGA)進行求解。最后通過案例分析驗證了本文方法較先前的方法對企業(yè)級業(yè)務組件識別具有更好的適用性和高效性。
[Abstract]:The business architecture integration of military information systems based on components is an important research content in military field, and the identification of enterprise business components is a key issue in business architecture integration.The current methods of business component identification are mostly focused on the software level business components, ignoring the enterprise factors such as organization and resources, while the current methods of business component identification have been proved to be very inefficient.In this paper, a new method of business component identification is proposed, which takes into account five design principles: cohesion, coupling, granularity, maintainability and reusability.Firstly, the business component is defined and formalized based on the business component model and the DoD AF(Department of Defense Architecture framework model. In order to analyze the business component quantitatively, the business model is transformed into a crud model, and six kinds of indexes are proposed.Then, the problem of business component identification is transformed into a multi-objective optimization problem, and simulated annealing hybrid genetic algorithm is used to solve the problem.Finally, a case study shows that the proposed method is more applicable and efficient than the previous method in identifying enterprise business components.
【作者單位】： Science
【基金】：Project supported by the National Natural Science Foundation of China(No.71571189)
【分類號】：F272;TP18

【相似文獻】

相關(guān)期刊論文 前1條

1 何萍;徐曉華;陸林;陳];;雙層隨機游走半監(jiān)督聚類[J];軟件學報;2014年05期

，

本文編號：1740200