Agent-Oriented Intelligent Distributed Computing And Its Applications

This dissertation primarily introduces the method to construct intelligent distributed systems utilizing agent technology, and we built an intelligent distributed geographic information services platform and an intelligent optimization-computing platform with some related applications based on the research theory result.Internet-based distributed computing mode will prevails in the future. With the development of Internet, the environment of distributed systems is very different from that of traditional ones. Due to the diversity and heterogeneity in terms of the types of client devices and network connections that people use to access the systems as well the special needs and preferences that end users might have, system adaptability, interoperability, reusability and reliability, etc are very complex and difficult to solve. Currently available solutions for building distributed systems have little abilities to face those challenges, new methods and technologies have to be chosen to address them. Agent technology is a new method for distributed computing, Agent is the big-granularity, high-intelligence and autonomy computing entity in distributed environments, and it has autonomy, interaction, reactivity characteristics. One side agent technology provides new approach for solving problems in constructing new distributed applications on Internet; on the other side it provides a reasonable conceptual model for deeply research on the characteristics of distributed systems. It is an effective way to build intelligent distributed systems which can solve the problems faced while building distributed systems currently.First we fully analyze and discuss the new problems and requirements of distributed computing on Internet-based mode, then a new agent-oriented framework for intelligent distributed computing is provided. We construct a layered model of the framework, introduce various agents into each layer and describe the function and components of each layer in detail, and then we design and define an agent communication language X-ACL, which is based on XML and SOAP protocol in Internert-computing mode.Through the agent-oriented framework for distributed computing and our analysis of the characteristics in distributed Geographic information sevices, we introduce agent technology into distributed GIS and build an agent-oriented service model for it. New spatial data model and spatial meta-data model are designed aiming not only for the data share and services share, but also for the collaboration between different GIS applications and Intelligent GIS services. We apply the research theory result into building a 863 High-tech project "Land Digital Spatial Data Uniformly Retrieving, Database Construction and Its Important Applications in Hangzhou ", and it is akenerl-part of "Digital HangZhou".Through the agent-oriented framework for distributed computing, we apply agent technology into building distributed intelligent optimization computing platform. A new approach for the management of resources in distributed computing is proposed. This platform can become a basis platform for the intelligent optimization application in industry, and it can provide high-performance computing services, flexible and effective development services.Addressing the real-project problems in Power Plant, we propose two rapid intelligent optimization algorithms for unit-commitment problem and coal-blending problem, and then build a distributed intelligent optimization system for a power plant in GuangDong.