Mixed model‐based and signal‐based approach for open‐switches fault diagnostic in sensorless speed vector controlled induction motor drive using sliding mode observer

Abstract

This study deals with a mixed model-based and signal-based approach for sensorless speed-controlled induction motor drive (IM) and insulated-gate bipolar transistors open-switch fault diagnosis. In comparison to the classical sensored drive systems, this structure presents particular characteristics in post-fault operation mode which can be taken into account for fault detection and identification process designing. Firstly, the fault effects analysis is achieved in the abc frame of the IM. Based on such analysis, a diagnostic algorithm, using the measured and estimated currents, is used to define the fault indices, which allow the detection of single open-switch, multiple open-switch and open-phase faults. The sensorless control algorithm used for rotor speed estimation, as well as the fault diagnostic algorithm, are all based on a first-order sliding mode observer. In addition to the simplicity and the fast fault detection, there are no-additional sensors or extra-hardware used by the proposed method. Experimental results, based on a dSPACE DS1104 controller board and a 3-kW induction machine, are shown to validate the proposed strategy.

Ensuring a Reliable Operation of Two-Level IGBT-Based Power Converters: A Review of Monitoring and Fault-Tolerant Approaches

“…b: STATE-SPACE MODEL [101]- [106] As the state model is application-dependent, the two-level voltage source inverter fed induction machine drive system (see Fig. 26) is taken as an example for illustration purpose.…”

“…For the fault isolation, i.e., to identify the faulty switch, the direction of the threshold is utilized. Similar to the Luenberger observer, the first-order sliding mode observer can also be utilized with the form of (34) [106].ẋ = Ax + Bu + I s (34) where represents the observer gain and I s is the switching vector.…”

Broken rotor bar fault detection using Hilbert transform and neural networks applied to direct torque control of induction motor drive

“…To detect the fault in the machine, many signal processing techniques are investigated. The first technique is focused on temperature, noise, vibration evaluation [11, 12]. This method is expensive, and the installation is very sensitive to noise.…”

Improved double‐surface sliding mode observer for flux and speed estimation of induction motors

“…The rotor field orientation is the most common method of vector control of IMs since this method decouples d and q axes current components at the steady-state situation (not transient situation); hence, the currents could easily follow their command values. The use of the current control method based on proportional-integral (PI) controllers with or without the use of dynamic decoupling terms is prevalent [1][2][3][4][5][6][7][8][9][10].…”

“…x˙= Ax + Bu y = Cx (1) If the supposed m-dimensional state vector x is measurable and matrices A, B and C are time invariant, design of SMO is straightforward. By adopting the general method detailed in [14], the designed observer has m sliding variables, and the movement of sliding mode is applied to the intersection of sliding variables.…”