基于案例推理和神经元网络的桥梁知识的复用  

Reuse of Bridge Knowledge Based on Case Reasoning and Neural Network

作　　者：彭卫兵[1] 任爱珠[1] 徐峰[1] 

机构地区：[1]清华大学土木工程系,北京100084

出　　处：《计算机工程》2007年第23期202-204,207,共4页Computer Engineering

基　　金：交通部西部交通建设科技基金资助项目(200331822314)

摘　　要：采用面向对象方法进行桥梁知识数据组织,建立了包含几何信息和非几何信息的桥梁案例数据,并通过系统建模平台,建立了和中小跨径桥梁通用图库中项目图纸信息一一对应的可供推理的案例数据库。案例数据采用SQL Server 2005数据库技术进行结构化存储。研究了适合桥梁结构设计的构件相似性定义,建立了获取桥梁结构设计相似案例方法。采用人工神经元网络技术,挖掘相似案例中蕴涵的知识,实现对最相似案例的修改,形成符合新的设计要求的桥梁设计案例。实现桥梁数据的参数和知识驱动,形成了基于挖掘以往设计案例中包含的知识的桥梁结构设计方法,为桥梁结构设计企业重复利用以往设计案例提供了有效的解决方法和新的思路。An available method of converting standard drawings into reusable bridge design case is developed. With this method, data object of design knowledge, which integrates both geometric descriptions and non-geometric description, is established with object-oriented technology. At the same time, bridge data are divided into several levels for better comprehensibility and lower programming complication. Design case is stored into database with SQL server 2005. Similarity definition of bridge case is carried out by analyzing bridge design knowledge in order to get relative similar design cases. Case intelligent adaptation technology is created by neural network for new bridge design, and so a bridge model, driven with knowledge and parameters, is achieved.

关 键 词：桥梁结构设计 知识复用 案例推理 人工神经元网络 

分 类 号：TU2[建筑科学—建筑设计及理论]