Font Size: a A A

Research On Support Vector Machine-Decision Tree Arithmetic And It's Application

Posted on:2008-09-18Degree:MasterType:Thesis
Country:ChinaCandidate:J M HanFull Text:PDF
GTID:2189360212476416Subject:Mechanical design and theory
Abstract/Summary:PDF Full Text Request
The extraordinary growth of technologies and market competition has already had a significant impact on commercial manufacture. There are some new characteristics, such as hugeness, distribution, high speed, automation and complexity. And these equipments lie on the very key position in pillar industries, such as energy industry, petrochemical industry, metallurgical industry, etc.When faults would happen, production efficiency falls or machines halts. Once catastrophic accidents would occur, the loss of property and life must be a disaster. Furthermore, each enterprise has to cut fault time and fault frequency down and prolong efficient life cycle in order to strengthen their marketing competition. Consequently, it is the primary task that enterprise has to face and take measures for fault diagnostic. On-line monitor and diagnosis system is designed to find out all kinds of hidden troubles in the running machines as soon as possible to prevent disaster from happening.Intelligent diagnosis technology stands for the development direction of diagnosis technology, and it is associated with artificial intelligent technology and provides the possibilities of fault diagnosis intelligentization. However, the traditional artificial intelligent methods based on empirical risk minimization have worse training generalization especially in limited samples. Statistical learning theory is based on the theory foundation and provides a uniform framework for learning subject of limited samples. Support vector machines(SVM) is a machine-learning algorithm based on statistical learning theory. This algorithm accomplishes the structural risk minimization principle. The fine performance of Support Vector Machines to limited samples attracts attention of investigations in fault diagnosis field. Fault diagnosis is a limited samples subject. The most predominance of SVM is a proper for limited samples decisions. The nature of the algorithm is acquiring connotative class...
Keywords/Search Tags:Fault diagnosis, Support Vector Machines(SVM), Sequential Minimal Optimization(SMO), SVM Decision Tree(SVMDT), DCOM, ActiveX
PDF Full Text Request
Related items