论文部分内容阅读
为利用神经网络实现桁架结构的形态映射机制 ,采用形态基因方法对桁架结构拓扑形态进行了数字化信息提取 .以桁架拓扑结构形态基因的方式表述桁架的拓扑结构 ,应用预先布置节点图、桁架特征基因串、基因分片、压缩等方法 ,将桁架的拓扑结构表述成一串 0～ 1之间的小数形式 ,使其代表桁架拓扑结构的形态特征 .结果将形态化的桁架结构转化成为数字化的信息 ,形成了规范化的神经网络数据输入方法 ,为进一步实现桁架结构拓扑形态推理作好了准备 .算例证明此方法有效 ,对桁架结构拓扑形态信息进行了有效的提取和压缩
In order to use the neural network to realize the morphological mapping mechanism of truss structure, the morphological genetic method was used to extract the topological information of the truss structure.The topological structure of the truss was expressed by the morphological genes of truss topology, String, gene fragment, compression and so on, the truss topology is expressed as a series of decimal forms between 0 and 1, which makes it represent the morphological characteristics of the truss topology.Results The morphological truss structure is converted into digital information, A normalized neural network data input method is formed, which is well prepared for the further realization of truss structure topological reasoning.Examples show that this method is effective, and the truss structure topological information is effectively extracted and compressed