基于耦合人工神经网络的工业材料选择与决策方法研究

为了实现工业材料的快速选择和决策,提出利用以往工业产品设计方案实例来训练神经网络.通过目标产品的编码信息和索引模式实现知识推理,从而获得所需的信息.文中首先研究了工业材料选择与决策方法的耦合神经体系结构,通过ART1网络的学习规则和工作规则,解决工业材料的选择和决策问题,并给出了ART1神经网络模式的链式32位二进制工业材料编码方法.通过实例推理、神经元网络、面向对象等技术,使得基于耦合神经网络的工业材料和决策系统能够有效地实现产品的方案设计,达到了预期目的.

Neural Network Based Selection and Decision of Industrial Materials

In order to achieve the rapid selection and decision a method for training the neural network with past industrial design program. Knowledge reasoning is achieved through the target product of the encoding information and the indexing mode, thus obtaining the necessary information. The paper studies first the coupled training the neural network for the selection and decision of industrial materials and then by using learning rules and working rules of the ART1 network, the selection and decision of industrial materials is achieved. Besides, implementation process of the ART1 network learning algorithm is given. Case-based reasoning, neural network and object-oriented methods enable the selection and decision system of industrial materials which is based on coupled neural network to effectively implement the design of prodcuts with the expected results obtained, process of the ART1 network learning algorithm is given. Case-based reasoning, neural network and object-oriented methods enable the selection and decision system of industrial materials which is based on coupled neural network to effectively implement the design of prodcuts with the expected results obtained.