Research On Agent-Oriented Software Analysis And Design Modeling Methodology And Environment Based On Grid

"Grid" is a hardware and software infrastructure that provides dependable, consistent, pervasive, and inexpensive access to high-end computational capabilities. Grid coordinates resource sharing and problem solving in dynamic, multi-institutional virtual organizations. Grid-based Problem Solving is called grid computing, which has emerged as a research hot spot field currently. Distributed environment based grid computing has characteristics such as heterogeneity, extensibility, openness and adaptability. Therefore, software systems based on grid infrastructure are different from software systems based on conventional distributed environment. It is to appear autonomous, cooperately, open and interoperable. We called this software form Grid Software, or Grid Distributed System.Progress in software modeling technologies over the past two decades has been made through the development of increasingly powerful and natural high-level abstractions with which to model and develop complex systems. Procedural abstraction, abstract data types, and, most recently, objects and components are all examples of such abstractions. Object has become the chief paradigm in software engineering field, whereas, object-oriented modeling method is a static and close technology framework in essence. From the perspective of Grid Software, object-oriented methods are difficult to adapt well to the features of grid software. The development of grid computing has challenged conventional software architectures, methods, technologies and platforms.If grid computing is considered from a Distributed Artificial Intelligence (DAI) perspective, it is similar to a Multi-Agent System (MAS), since entities in a MAS and the grid have autonomous behavior. Autonomy is one of the key abstraction features of agents. Other features of agents relevant to grid entities include social ability, as well as adaptability and pro-activeness. Different In addition, software agents are heterogeneous, which could be implemented in different programming languages and technologies. These characteristics of agents justify the selection of agents as the main abstraction for modeling grid software.Most of today's Agent-oriented modeling methods emphasize modeling conventional close software architecture, and while model a grid software, they lack of effective concept abstraction and declarative models. Based on the research of grid architecture and main agent-oriented software modeling methods, this thesis puts forward a new modeling method for agent-oriented analysis and design based on OGSA architecture, AOMG(Agent-Oriented Modeling based on Grid). The main research contents of AOMG method include:1. Proposes a multi-agent architecture of the distributed software that based on OGSA grid infrastructure.