DAF-Net and multi-agent based integration approach for heterogeneous CIM information systems
Abstract
The objective of this research is to investigate an integrated architecture for heterogeneous CIM information systems. Its features are: (1) automatic coordination of interdependent data activities; (2) provision of high-quality data to users, and (3) integration by demand. To achieve these features, two research tasks are performed: development of the coordination model of data activities, and of the integration model of heterogeneous information systems. For concurrently coordinating data activities, an advanced workflow model called Data Activity Flow Net (DAF-Net) is developed. To construct DAF-Net systematically, a three-phase building method is proposed. By the DAF-Net model, the automation of CIM data activities and the provision of high-quality data to users can be realized. For integrating heterogeneous information systems, a multi-agent based integration model, called Agent-based Integration Model of Information Systems (AIMIS), is developed. AIMIS consists of global coordination agents, local coordination agents, and an agent interaction protocol. Two agent interaction protocols are developed. By the modularity modula of local coordination agents, the concept of integration by demand can be accomplished. Two sets of experiments were performed to investigate DAF-Net and AIMIS design and effectiveness. Parallel Virtual Machine (PVM) software and a MRPII scenario were employed to conduct the experiments. The main findings are: (1) DAF-Net with AIMIS can preserve data consistency more tightly compared with a periodic update without DAF-Net, and (2) simple (decomposite) DAF-Net and an agent interaction protocol with learning capability are relatively more resilient to information system failures. In addition, PVM has been proved to be a highly effective tool for these experiments.
Degree
Ph.D.
Advisors
Nof, Purdue University.
Subject Area
Industrial engineering|Systems design
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.