Statistical and Fusion Based Hybrid Approach for Fault Signal Classification in Electromechanical System
Motor fault diagnostics in dynamic condition is a typical multi-sensor fusion problem. It involves the use of multi-sensor information such as vibration, sound, current, voltage and temperature, to detect and identify motor faults. According to our experiments in BLDC motor controller results, the system has potential to serve as an intelligent fault diagnosis system in other hard real time system application. To make the system more robust we make the controller more adaptive that give the system response more reliable by the multisensory fusion techniques. We introduce a hybrid model based new methods and evaluate the performance of the proposed information fusion system. Finally, we report the efficiency of this system in dealing with controller stabitility and its nonlinear information that may arise among the sensors.
KeywordsMotor diagnosis Information fusion Sensor fusion Support Vector Machine Sort Term Fourier Transform Brash less Direct Current Motor signal classification Fault Classifier
Unable to display preview. Download preview PDF.
- 1.Runkler, T., Sturm, M., Hellendoorn, H.: Model based sensor fusion with fuzzy clustering. In: The 1998 IEEE International Conference on Fuzzy Systems Proceedings, IEEE World Congress on Computational Intelligence, vol. 2, pp. 1377–1382 (May 1998)Google Scholar
- 2.Luo, R., Kay, M.: A tutorial on multisensor integration and fusion. In: 16th Annual Conference of IEEE Industrial Electronics Society, IECON 1990, pp. 707–722 (November 1990)Google Scholar
- 4.Lee, C., Xu, Y.: Theoretical study on a new multi-sensor system. In: Proceedings of the First ISA/IEEE Conference Sensor for Industry, pp. 187–191 (November 2001)Google Scholar
- 5.Abderahman, M., Kandasamy, P.: Integration of multiple sensor fusion in controller design. In: Proceedings of the American Control Conference, vol. 4, pp. 2609–2614 (May 2002)Google Scholar
- 6.Banerjee, T.P., Das, S., Roychoudhury, J., Abraham, A.: Implementation of a New Hybrid Methodology for Fault Signal Classification Using Short -Time Fourier Transform and Support Vector Machines. Advances of Soft computing, pp. 219–225. Springer, Heidelberg, ISBN 978-3-642-13160-8Google Scholar
- 10.Vemuri, A., Polycarpou, M.: On the use of on-line approximations for sensor fault diagnosis. In: Proceedings of the American Control Conference, vol. 5, pp. 2857–2861 (June 1998)Google Scholar
- 11.Durrant-Whyte, H.: Elements of sensor fusion. IEE Colloquium on Intelligent Control, 5/1–15/2 (1991)Google Scholar
- 13.Mall, R.: Real time systems. Theory and practice. Pearson Publication (2007)Google Scholar