Current-sensor fault detection and isolation for induction-motor drives using a geometric approach. (August 2016)

• Record Type: Journal Article
• Authors: Aguilera, F.; de la Barrera, P.M.; De Angelo, C.H.; Espinoza Trejo, D.R.
• Is Part Of: Control engineering practice. Volume 53(2016)
• Page Start: 35
• Page End: 46
• Publication Date: 2016-08
• Subjects: Induction motor drive -- Fault detection and isolation -- Nonlinear observer -- Sensor -- Geometric approach; Automatic control -- Periodicals; 629.89
• DOI: 10.1016/j.conengprac.2016.04.014
• Languages: English
• ISSNs: 0967-0661

Abstract

Abstract: This work presents the design of a current-sensor fault detection and isolation system for induction-motor drives. A differential geometric approach is addressed to determine if faults can be detected and isolated in drives with two line current sensors by using a model based strategy. A set of subsystems is obtained based on the observability co-distribution, whose outputs are decoupled from the load torque (detectability) and only affected by one of the sensors (isolability). A bank of observers is designed for these subsystems in order to obtain residuals for the fault detection and isolation. It is demonstrated that the proposed strategy allows detecting single and multiple sensor faults, including disconnection, offset and gain faults. Experimental results validate the proposal. Abstract : Highlights: A geometric approach for sensor-fault detection is applied to induction motor drives. Single and multiple sensor fault isolation is obtained without physical redundancy. Residuals sensitive to disconnection, offset and gain faults were obtained. Proposed system is able to detect the recovery of a sensor from a faulty condition. Fast fault detection and isolation of fault is obtained.