基于神经网络学习机制的应急决策支持中间件模型

Middleware Model of Emergency Decision Support System Based On Neural Network Learning Mechanism

在传统的决策支持系统模型库、知识库、数据库、解释器等组成部件的基础上,引入中间件技术,构建基于神经网络学习机制的应急决策支持中间件模型(MMEDSS)。该模型由解释中间件(EM)、数据库中间件(DBM)、数据挖掘中间件(DMM)、模型库中间件(MBM)、知识库中间件(KBM)组成,其中EM负责和用户交互,DBM负责数据库访问,MBM和KBM负责向EM提供模型和规则,DMM应用神经网络学习算法,通过案例学习动态更新模型库。

The traditional decision support system （DSS） comprises model base, knowledge base, database and etc. On this basis, this paper proposes a middleware model of emergency decision support system （MMEDSS） based on neural network learning mechanism by referring to middleware technology and neural network technology. This model consists of explaining middleware （EM）, data base middleware （DBM）, data mining middleware （DMM）, model base middleware （MBM） and knowledge base middleware （KBM）. Each middleware interacts with another via object request broker （ORB）. Here EM is responsible for interaction with user; DBM takes charge for database access; MBM and KBM are responsible for maintenance of model base and knowledge base as well as the supply of models and rules for EM. DMM answers the dynamic update of model base by data mining and case learning with artificial neural network （ANN） algorithm.