Discovery of Faults in Electrical Machines Using Data Mining Tools  
 

Professor Sérgio Pinheiro dos Santos
Adaptive System Laboratory – Department of Electrical Engineering
Federal University of Rio Grande do Norte,
Natal-RN,
Brazil
E-mail: sergio_p_santos@hotmail.com
 

Abstract: Data mining is predicted to be one of ten emerging technologies that will change the world. According to the Gartner Group, “Data mining is the process of discovering meaningful new correlations, patterns and trends by sifting through large amounts of data stored in repositories, using pattern recognition technologies as well as statistical and mathematical techniques.” With the rapid development of devices and computer technology, more and more on line systems have been installed to monitor the running conditions of machines groups, which resulted in the massive quantities of accumulative data. Now, many large-scale databases and data warehouses have come into being, some even arrived at the level of TB. And many researchers are faced with the urgent problem, as how make full use of such data, and whether it is feasible to extract valuable knowledge and condition monitoring. Maintenance costs represent a significant percent of industrial products’ cost. The induction motor (IM) is the most commonly used type of ac motor. Examples are found in pumps, refrigerator compressors and so on. IMs are vulnerable problems as temperature, undesirable vibrations, unbalance of stator currents and broken bars, usually detected when the equipment is already broken, and sometimes, with irreversible damages. Data mining is a powerful technology with great potential to help researchers in faulty discovering. With the development of artificial intelligence and database technology, some machine learning algorithms including case-based learning, decision trees, genetic algorithm and pattern recognition tools are available in application now. In addition, advanced microprocessor manufacturing technology makes it possible to support the large-scale data analysis and processing. In this plenary speaker, it is outlined the basic notions in this area, define some key ideas and problems, and motivate their importance. Different data mining tools are explained as well some applications in real world problems.