Development of intelligent drive systems using embedded condition monitoring and fault diagnosis algorithms

Nowadays, when the rapidly developing industries are turned into computer-based automation, the realization of real-time condition monitoring and fault diagnosis systems is very important. There are various researches on condition monitoring and fault diagnosis algorithms that can work integrated with drive systems in industrial enterprises. Since pre-determination of fault that can occur in induction motors in industrial systems will provide the potential production continuity, it is very important to develop a predictive maintenance method that includes condition monitoring and fault diagnosis algorithms in induction motors. In this thesis study, an embedded application is developed for condition monitoring and fault diagnosis algorithms and intelligent driver systems using these algorithms. For this purpose in extent of the thesis, firstly a fault diagnosis approach based on time series analysis is presented with experimental results. For this purpose, it is given the algorithms developed for inverter-fed faults in induction motors. In the other work, an algorithm that uses the Hilbert transform for the detection of eccentricity faults is proposed. In this study, the performance of the proposed approach is presented by experimental results. At the last stage of the thesis, a design has been realized for implementing an embedded driver system with the ARM-based STM32F746 kit of developed algorithm and the results are given. As a result, in this thesis study is performing suitable for real-time work a condition monitoring and fault diagnosis algorithm and these algorithms have been confirmation on an embedded system. This thesis study was supported by SAN-TEZ project with code 0692.STZ.2014.