Maximization and Knowledge Description of Experimental Information

实验在结构工程研究中起着非常重要的作用，它是证明我们所建立的理论的有效性和使我们了解结构及其构件的真实行为的有效手段。沿着这条路线进行了许多试验，几十年来积累了有用的发现。近年来，由于实验所用硬件和软件的发展，测试的种类增加，规模显著扩大。因此，最近的测试提供的信息被认为比以前更丰富和复杂。在这里，出现了一个看似非常严重的问题，即我们如何与具有相同兴趣的其他机构分享如此丰富而复杂的信息。当然，许多研究人员，无论是当代的还是未来的，都希望参考与他们的研究相关的实验信息。然而，除非他们所需要的信息类型能够被轻松地获取，否则这些信息无论多么重要，都无法发挥其主要作用。这篇题为“实验信息的最大化和知识描述”的论文报告了一个项目，该项目发明了一个原型，用于有效存储宝贵的实验信息，并根据希望参考它的人的要求提供它。本文为获取钢梁柱强度和变形能力的试验资料，研制了原型机。建立了钢梁柱强度和变形能力试验结果的关系型数据库，并设计了方便的图形显示程序，以便及时获取和评估数据。利用该原型数据库对设计规范中提供的强度公式的精度进行了标定，结果表明，该数据库对于量化设计规范中强度公式的精度和变化程度是非常有效的。数值敏感性分析为我们提供了各种参数对钢梁柱强度和变形能力的影响程度的信息。所得结果用于检验试验结果与分析结果之间的相关性。我们发现，这种审查反过来又引导我们对数据库的进一步应用获得新的想法。为了方便测试数据和分析数据之间的信息交换，分析得到的结果以生成规则的格式进行描述和存储。少

Experiment plays a very important role in the research of structural engineering, serving as a useful means to prove the validity of theories we have developed and also to make us aware of the true behavior of structures and their components. Numerous tests have been carried out along this line, and useful findings have been accumulated over many decades. In recent years, the variety of tests increases, and their scales are enlarged significantly, thanks to the development of both hardware and software employed in the experiment. Accordingly, the information provided by recent tests is deemed much richer and complicated than before. Here, a concern, seemingly very serious, arises as to how we can share such rich as well as complicated information with other bodies who have the same interests. Certainly, many researchers, both contemporary and future, should like to refer to experimental information related to their studies. However, unless the type of information that they need can be … More accessed without much troubles, the information, no matter how important it may be, cannot achieve its major role.This paper, entitled, "Maximization and Knowledge Description of Experimental Information," reports on a project, in which a prototype was invented for the procedures to effectively store precious experimental information and make it available upon request by those who wish to refer it. In this paper, the prototype was developed for the experimental information on the strength and deformation capacity of steel beam-columns. A relational database that contains test results regarding the strength and deformation capacity of steel beam-columns was developed, and a program for handy graphic presentation was devised for prompt acquisition and evaluation of the data. Using this prototype database, accuracy of strength formulas provided in design specifications was calibrated, and it was demonstrated that such a database is extremely effective to quantify the accuracy as well as the degree of variation of those formulas. Numerical sensitivity analysis was also carried out, providing us with information on the degree of effects of various parameters on the strength and deformation capacity of steel beam-columns. The results obtained were used to examine the correlation between the test and analytical results. The examination, in turn, was found to lead us to obtaining new ideas on the further application of the database. To facilitate the interchange in information between the test and analytical data, the results obtained from the analysis were described and stored in the format of production rule. Less