基于贝叶斯推理的浇注系统选型方法研究  

Research on model selection method for gating system based on bayesian inference

作　　者：刘德仿[1] 谭桂琴[2] 王斌[1] 吴宏攀[3] 

机构地区：[1]盐城工学院优集学院,盐城224001 [2]南通大学机械工程学院,南通226019 [3]江苏大学汽车与交通工程学院,镇江212013

出　　处：《机械设计与制造》2011年第6期46-48,共3页Machinery Design & Manufacture

基　　金：江苏省科技攻关项目(BE2007054);盐城市科技发展计划(YK2008001;YK2008091)

摘　　要：针对压铸模浇注系统设计过程中存在大量不精确数据的情况,提出基于贝叶斯推理算法的压铸模浇口类型选择方法。结合浇注系统的类型划分,建立相应的类型标准模型以及典型属性模糊向量,并采用模糊数学的方法处理设计信息中不精确约束值,最后将贝叶斯理论应用于推理过程中。该方法将概率逻辑与模糊数据处理的思想结合在一起,将贝叶斯推理方法应用于基于智能化设计的浇注排溢系统中,提高了对设计信息的获取能力以及产品设计的智能化程度。Concerning a great quantity of inaccurate data available in the design process for gating system of die-casting dies,a pouring gate selection method for the gating system of the die-casting die based on bayesian theory is proposed,which constructs a standard model for relevant type and a fuzzy vector with typical nature combing the type division of the gating system.In this case,the inaccurate constraint value in design could be processed with fuzzy mathematics,and finally the bayesian theory shall be applied in the course of reasoning.The method applies Bayesian Inference into the pouring and exhausting system based on intelligentized design,and combines probability logic with the idea of fuzzy data processing to improve the ability of obtaining design information and the intelligentization of product design.

关 键 词：贝叶斯推理 类型选择 不精确信息 

分 类 号：TH16[机械工程—机械制造及自动化] TP182[自动化与计算机技术—控制理论与控制工程]