Expertní systém pro identifikaci vad odlitků

Abstract

The computer technologies have been become the moving by force human community already from last decade of past century. They intervene in to many science branchies, they touch business area, they influence our personal life more and more. Their meaning shoots up in the foundry technologies too, where the various programmes make use of the area of statistic drive quality control of castings, they make use of the following of significant technological parameters, in primary row a lot of information systems shoot up by means of that it´s possible effectively to solve problems of the identification casting defects. These programmes are called the knowledge-based systems, they are based on the knowledges databases, which draw from the knowledges of human experts to working in foundries. The Expert system for identification casting defects is introduced in this paper, this system identifies defects according to their external symptoms on the basis visual classification. This programme was called ESVOD, it was generated in the information system Visual FoxPro and can be used in common personal computer of foundry users. Higher form solving of identification casting defects by the help of information technology is using of artifical neuronal nets. Neuronal experts systems as modern element of information technologies continue to be more often used as predictor and optimalization tools for the following foundry processes. The using of neuronal artificial nets finds always wider exercise not only in the area foundry industry, but also in other metallurgic industry.