PhD Defense: Xin Xue
Ph.D. Dissertation Defense Health Monitoring of Drive Connected Three-Phase Induction Motors from Wired Towards Wireless Sensor Networks
Xin Xue, Ph.D. Candidate
Advisor: Professor V. Sundararajan
Wireless sensor network (WSN), one of the featured technologies that the U.S. Department of Energy (DOE) has identified to help improve the overall energy efficiency of US industry, provides a potentially low-cost approach for the health monitoring and fault diagnosis of induction motors. The reduction of machine failures increases plant efficiency and productivity. Low-cost wireless sensor systems can help the health monitoring of manufacturing equipments by eliminating the cost of installation and increasing the flexibility of system diagnosis.
This research focuses on developing a nonintrusive, condition based health monitory system for drive connected induction motors using the wireless sensor network method. A hierarchical classification system is designed for motor fault diagnosis. To simulate and analyze a wide range of fault conditions that may arise in induction motors, an experimental test bed is also developed. Three major branches of induction motor faults are studied, either individually or in combination. Wired sensors are first used to find optimal features for motor fault classification. After performing feasibility studies of wireless sensors in electric machinery, two wireless sensor nodes are developed and implemented in the motor health monitoring and fault diagnosis system. The experimental results demonstrate the effectiveness and generalizability of the wireless sensor system for motor health monitoring and fault classification.