This research focuses on the application of the extended Kalman filter (EKF) in vector-controlled induction motor drive, designed to ensure tolerance against various types of current sensor (CS) faults. The EKF-estimated stator currents are initially used in a residual-based detector. In the event of a current sensor fault, these estimated currents replace the faulty measurements to compensate for the failure. To enhance current estimation accuracy, the EKF also estimates the coefficient 'd', which compensates for variations in both motor resistances. Simulation studies validate the effectiveness of the proposed approach in handling different types of CS faults, including gain changes, offsets, saturation, and complete signal loss, under various operating conditions, including regenerative braking.
The data presented in the attached set are the results of simulation studies of a detection and compensation system for stator current sensor faults based on the Extended Kalman Filter (EKF), dedicated to a vector-controlled induction motor drive,. described in details in the mentioned publication (M. Miniach, T. Orlowska-Kowalska, Adamczyk M., Application of the extended Kalman filter in current sensors fault-tolerant control of the induction motor drive, 2024 IEEE 21st International Power Electronics and Motion Control Conference (PEMC); 30 September 2024 - 03 October 2024 , Pilsen, Czech Republic, https://doi.org/10.1109/PEMC61721.2024.10726418).
The simulations were carried out in the MATLAB/Simulink environment.
The considered system includes two current sensors (on phase A and B), each of which may be subjected to the following faults:
OC – complete signal loss; G – gain error; OFF – occurrence of a DC component; SAT – saturation.
File Description:
1) Folders:
"m100_A-OC_B-G07" contains measurements for:
operation under a load equal to 100% of the rated torque,
phase A sensor fault type: OC,
phase B sensor fault type: G (actual current scaled by a factor of 0.7).
"m30_A-OFF02_B-SAT03" contains measurements for:
operation under a load equal to 30% of the rated torque,
phase A sensor fault type: OFF (signal offset by +0.2),
phase B sensor fault type: SAT (signal value limited to ±0.3).
2) Included Signals (all expressed in per unit, p.u.):
d_ref, d_EKF – coefficient "d" values: reference and EKF-estimated, respectively,
isA_IM, isB_IM – actual phase currents,
isA_faulty, isB_faulty – phase currents measured by the faulty sensor,
isA_EKF, isB_EKF – phase currents estimated by EKF,
lambda_detektor, lambda_real – fault coefficient: defined by the detector and reference, respectively [1-no fault, 2-sensor A faulty, 3-sensor B faulty, 4-both sensors broken];
w_ref, w_mea – reference and measured speed, respectively,
t_em – electromagnetic torque,
psira_IM, psirb_IM – actual alpha-beta components of the rotor flux,
psira_EKF, psirb_EKF – alpha-beta components of the rotor flux estimated by EKF.
(2025-02-18)
