水闸老化评估神经网络专家系统知识获取方法

专家系统的工作过程是一个获得并应用知识的过程,恰当的知识表示对于专家系统的构建具有重要意义.以上海地区多座中小型水闸老化检测实例为基础,开发出水闸老化评估神经网络专家系统ESANN-SLUICE.此系统从实例中生成数据,再对神经网络进行训练和管理,从而建立权值矩阵,然后转换成相当规则的知识库,提供给专家系统使用.该方法是水闸老化评估知识获取的一种新方法,是对传统知识获取方式的拓展和补充.最后通过实例证明该方法是可行和有效的.

Knowledge Acquisition in ANN Expert System of Sluice Aging Assessment

Operation of an expert system(ES) is basically a process of knowledge acquisition and application,in which a proper knowledge presentation is important in the ES formation.The aging assessment system ESANN-SLUICE is developed based on many aging assessments of medium and small-sized sluice in Shanghai area.Data used to train and manage the ANN is obtained from the assessment.A weighting matrix is then established and transformed into a knowledge base.This is a novel method that supplement to the traditional knowledge acquisition.An example is given to demonstrate feasibility and effectiveness of the method.