Font Size: a A A

Development Of Artificial Intelligence Design System Of Magnesium Alloys

Posted on:2008-06-20Degree:MasterType:Thesis
Country:ChinaCandidate:H J WangFull Text:PDF
GTID:2121360215997681Subject:Mechanical design and theory
Abstract/Summary:PDF Full Text Request
Study on Magnesium alloys is one of hottest research fields in recent years. Material design is an advanced study in material science. Artificial intelligence design system of Magnesium alloys, which is a practical application of artificial neural network in material design, has been developed to accelerate the development of Magnesium alloys'study in China.Basic theory of material design and artificial neural network are discussed to prove the necessity and positive value of developing this system. According to the need analysis, population structure was designed and realization method of this system was selected. At present, there are three parts in this system, includeing Magnesium alloys database, RBF network and man-machine interface. Microsoft SQL Server 2000, JBuilder 9 and Matlab 7.1 are adopted as development implements.In algorithmic study, firstly, the basal theory of RBF network is analysed incisively. Secondly, newrb function, K-means value cluster-algorithm, GAP-algorithm, CVS-algorithm and Genetic-algorithm which are used to ascertain the network's parameters are discussed. Thirdly, specifically for RBF network's parameters'character, five optimal algorithms are proposed, which are combined with some algorithms'good points.In the aspect of system realization, this system is developed with various softwares. Especially, JMatLink is adopted to realize Java calling programs written with Matlab directly, which fully utilizes Java's advantages and Matlab's powerful Data handling ability and plenty tool case functions.In this system, some potential and non-linear laws, such as the impact of alloys'compositive chemical elements over properties, are detected by calculating those experimental data because of artifical neural network's non-linear character. Artificial intelligence design of Magnesium alloys has been realized by predicting properties of designed materials. As a result, this system is helpful for new material design. Manpower, material resources and time can be retrenched. Meanwhile, this system is an application of material design in Magnesium alloys and has positive meaning to the research of material design.
Keywords/Search Tags:Magnesium Alloy, neural network, RBF, newrb, K-means, GAP, GA, JMatLink
PDF Full Text Request
Related items