基于SOA架构的平行管理信息系统设计

Design of parallel management information system based on service-oriented architecture

通过综合考虑企业生产过程中的工程复杂性和社会复杂性,平行管理为企业的综合管理和预测管理提供1种很有前景的理论方法,然而平行管理的软件技术架构体系需进一步完善。本文基于SOA技术,提出平行管理信息系统的概念、系统架构及其与现有信息系统的交互方式。结合乙烯生产系统,将平行管理信息系统分为协同层、应用层、服务层、组件层和资源层,在服务层和组件层构建和真实乙烯系统相一致的人工乙烯系统,在协同层和应用层通过流程编排和服务组合实现企业管理的计算和优化。最后以乙烯员工行为评估为例,说明了系统的构建过程。与传统的管理信息系统不同,平行管理信息系统将虚拟系统与实际信息系统互联互通,并融入了人的行为和管理制度等社会性因素统筹考虑。系统具有开放性、可灵活重构的特点,能快速响应企业业务变化和实验设计需求,为复杂生产系统的试验、培训、优化等多个管理目标提供了可行的技术实现手段。

Considering the engineering complexity and social complexity of the enterprise production, Parallel Management System （PMS） provides a promising approach for the comprehensive and the predictive management of the complex production system, however, the system architegture of the PMS need to be further perfected. In this paper, the concept of Parallel Management Information System （PMIS） from the PMS methodology and its system architecture are investigated in detail based on Service-Oriented Architecture （SOA）, as well as the parallel execution modes with the existing information systems. As an example, the PMIS for the complex ethylene production is divided into the collaborative layer, application layer, service layer, component layer and resource layer. The artificial ethylene systems alike the real-ethylene systems are constructed in the service layer and component layer, and the computational experiments and parallel execution are carried out in the collaborative layer and application layer by the process choreography and Service Composition. Finally, the design and the architecture of the PMIS are illustrated by one behavioral management computing example. Compare to traditional management information systems, PMIS combines the virtual system with the actual information system and takes social factors into consideration such as human behavior and management policy. PMIS is open and flexible reconfigurable, and can respond to changes in enterprise business, and then provides a feasible technical approach of experiment, training, optimization and other management objectives.