Structural Morphogenesis of Spatial Structure by Using Life-like Approach

机体具有良好的生命维持机制和信息传递机制功能。为了物种的繁荣，这些功能是有机体中不可思议的潜在力量，从原始时代到现在，通过各种方式获得。代表性的功能有：进化、遗传、学习、免疫、自治去中心化或自组织等。应用它们的功能的方法已成为结构形态发生研究的重点。一般来说，这种方法被称为类生命方法。逼真的方法是一种模仿这些机制的计算算法。在结构形态发生和结构优化方面，基于类生命方法的研究越来越多。遗传算法（GA）和神经网络（NN）是类生命方法中众所周知的计算方法，有许多应用实例。人们已经越来越认识到，这些分析程序是解决这些结构问题的重要手段之一。本文提出了一类自组织、遗传算法和免疫算法（IA）在类生命方法中的复合计算过程。该方法可应用于结构形态发生、结构优化和结构逆问题的分析模型。特别地，这个程序被应用于空间结构设计的结构形态发生。该过程的核心是基于有限元方法的节点邻域模型的元胞自动机操作。通过各种数值模型验证了该方法的有效性和有效性。此外，还提出了一种具有类生命方法的系统思想。文中提出了基于这一思想为试验目的而制作的实验系统。该系统是盆景形态发生和育种的一种方法。分析结果阐明了用类生命方法构造形态发生系统的可能性。少

The excellent functions for the life support mechanism and the information transfer mechanism are equipped in the organism. For the prosperity of the species, such functions are the marvelous potential powers in an organism, acquired using various means from primeval times to the present. As representative functions, the following are well known: evolution, heredity, learning, immunity, autonomous decentralization or self-organization and so on. The method for applying their functions has been spotlighted in the structural morphogenesis. Generally, such a method is called a life-like approach. The life-like approach is a computational algorithm imitating these mechanisms. In a structural morphogenesis and a structural optimization, the number of studies based on a life-like approach has been increasing. The genetic algorithms (GA) and the neural networks (NN) are computation procedures well known in the life-like approach, and there are many application examples. It is already being re … More cognized that these analysis procedures are one of the important means of solving those structural problems.In this report, the computation compound procedure for a self-organization, GA and Immune Algorithm (IA) classified into the life-like approaches is proposed. The proposed procedure can be applied to analysis models for a structural morphogenesis, a structural optimization and a structural inverse problem. Especially, this procedure is applied to structural morphogenesis for a spatial structure design. The kernel of our procedure is cellular automata (CA) manipulation using the node neighborhood model based on the finite element method (FEM). Effectiveness and the validity of the proposed method were examined with various numerical models. Moreover, an idea of system with the life-like approach is shown. It presented the experimental system made for trial purposes based on this idea. This system is the method of morphogenesis in the bonsai and breeding. The possibility of a system for structural morphogenesis with the life-like approach is clarified from analysis results. Less