Chaotic Extension Neural Network Theory-Based XXY Stage Collision Fault Detection Using a Single Accelerometer SensorReportar como inadecuado




Chaotic Extension Neural Network Theory-Based XXY Stage Collision Fault Detection Using a Single Accelerometer Sensor - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Department of Electrical Engineering, National Chin-Yi University of Technology, Taichung 41170, Taiwan





*

Author to whom correspondence should be addressed.



Abstract The collision fault detection of a XXY stage is proposed for the first time in this paper. The stage characteristic signals are extracted and imported into the master and slave chaos error systems by signal filtering from the vibratory magnitude of the stage. The trajectory diagram is made from the chaos synchronization dynamic error signals E1 and E2. The distance between characteristic positive and negative centers of gravity, as well as the maximum and minimum distances of trajectory diagram, are captured as the characteristics of fault recognition by observing the variation in various signal trajectory diagrams. The matter-element model of normal status and collision status is built by an extension neural network. The correlation grade of various fault statuses of the XXY stage was calculated for diagnosis. The dSPACE is used for real-time analysis of stage fault status with an accelerometer sensor. Three stage fault statuses are detected in this study, including normal status, Y collision fault and X collision fault. It is shown that the scheme can have at least 75% diagnosis rate for collision faults of the XXY stage. As a result, the fault diagnosis system can be implemented using just one sensor, and consequently the hardware cost is significantly reduced. View Full-Text

Keywords: master and slave chaos error systems; extension neural network; XXY stage; dSPACE master and slave chaos error systems; extension neural network; XXY stage; dSPACE





Autor: Chin-Tsung Hsieh, Her-Terng Yau * , Shang-Yi Wu and Huo-Cheng Lin

Fuente: http://mdpi.com/



DESCARGAR PDF




Documentos relacionados