Study On Decision Support System Based On Agent

The study on decision support systems has been a hot spot since the concept of decision support systems proposed in the 1970s. In recent years the development of new technologies and the improvement of decision-making requirements further promote the development of decision support system. Especially, the development of Agent technology which is originated from the field of Artificial Intelligence, provide a new research direction to enhance decision-making capability of the traditional decision support system. However, as Agent technology itself is still in a process of continuous improvement, the existing study on the framework of agent-based decision support system often tended to be concentrated in describing functional parts of the framework.But usually lack of theoretical foundation and software engineering methods. Such a framework is often poor in adaptability and scalability. The paper mainly done the following work to solve the problems which would be encountered in the process of integrates the Agent technology and decision support system.(1)An entire solution on agent-based decision support system which is guided by the framework of the decision support system which meets to agent standards (FIPA), based on Agent-Oriented software engineering method and combined with the relevant Agent development platform is proposed.(2) In this paper, taking the development process of a steel group's decision support systems as the background to refine and validate the proposed solution. Under the premise of complying with FIPA standards, according to the actual background in the project, it selects MaSE method and JADE development platform to complete the analysis, design and implementation of the agent-based decision support system.(3) The paper analysed and summarized the MaSE method. Improving the insufficient convergence between the step of creating agent classes and the step of constructing conversations by adding the step of constructing interaction model, it ensures the intergrity of conversations between agents.(4) The paper presents a scalable framework in which JADE combines with the J2EE. The combination of JADE and J2EE in the presentation layer has been carried out to ensure the scalability of the system.The operation of the agent-based decision support system which was implemented by this paper shows that, the system could provid strong support to the technologic route planning process and the production planning process.And it achieved a certain economic benefits. The achievement of operation further verified the feasibility, scalability and practicality of the solution which was proposed by the paper.