Design & Realization Of Material Dispatching Decision Support System Based On Case-based Reasoning

In the past 20 years, the application of Intelligent Decision Support System (IDSS) has made enormous progress and been widely used in domains of consultation, diagnosis, prediction, design, and management. And IDSS has got wider space with the continuous development of internet/intranet technology and flourish of the artificial intelligence technology. In order to solve the difficulties of knowledge expressing and knowledge reasoning in IDSS, an IDSS model, which combines case-based reasoning with Decision Support System (DSS), is proposed after introducing theories, working mechanism and existing problem of case-based reasoning (CBR). Subsequently the thesis focuses on case presentation and case retrieval, two key technologies in CBR. Representing cases with XML meets the need of the integration between different CBR systems and other application systems. Thus it is easy to realize information integration, distributed case store and search. Then a retrieval model of introspective learning is given to decrease the time of case retrieval, which can vary with the number of cases in case base obviously. Combining the author's research practice, the thesis analyzes the system structure of material dispatching in Longtan project in detail, and points out that scientific daily plan of Nandan is the crucial factor to ensure the progress of Longtan project. Then case-based reasoning is used to make daily dispatching plan of Nandan after the feasible demonstration of case-based reasoning which is applied in material dispatching of Longtan project. As an application research, operation processes have been analyzed with the thought of J2EE. Then a soft prototype system of DDPIDSS (Daily Dispatching Plan Intelligent Decision Support System), based on J2EE, is fulfilled.