승강기용 동기전동기 진단시스템 개발 및 SVM을 이용한 결함 분류

Abstract

Condition monitoring and diagnosis of the machinery have received considerable attention from the industry in order to improve the reliability of the operation and reduce production losses due to the malfunction or failure. The importance of those technologies is increasing more and more. Inspection for the safety of elevator is a check on a regular basis by the relevant authorities of the country, but the fundamental problem that affects the machinery failure and ride quality is usually caused by reducing costs of production, faulty design and improper installation. Maintenance worker who is in charge of these problem just checks kinds of magnetic contacts including switch, so noise and vibration problems caused by the root problems such as mechanical, electrical and electronic problem are difficult to deal with. In order to improve these structural problems, condition monitoring techniques have been introduced. Various researches to detect and to classify the defects of the induction motor and reducer have been conducted in the field of elevator industry and general industry. Recently, synchronous motors for elevator are widely used in low and medium speed elevator. However, research on noise and vibration characteristics of the synchronous motor have been limited to fault analysis of the electromagnetic force according to the electrical characteristics. Therefore, research on condition monitoring system and malfunction diagnosis using vibration signal of synchronous motor have not been conducted. Meanwhile, the existing condition monitoring equipment for elevator have just measured ride quality using micro electro-mechanical system (MEMS) sensor until now. This technique has disadvantage due to its difficulties to monitor vibration of traction machine perfectly. In addition, FFT analyzer for rotating machinery is also needed for monitoring the vibration of traction machine. So the equipment that can monitor vibration of traction machine and ride quality is needed in one system. In this thesis, a three-axis sensor system for measuring ride quality with excellent frequency response characteristic was developed. Using this, condition monitoring system to measure and analyze ride quality and malfunction of traction machine in elevator was developed, and verify its effectiveness. Additionally, the faults of synchronous motor can occur in the process of manufacturing. Through experimental works, the vibration signals were analyzed and classified. Artificial intelligence algorithm namely support vector machine (SVM) was implemented and trained using acquired signal, then automatic diagnostic system was configured and its effectiveness was verified.