0006EENG
0006EENG
ae 21
Impact
Factor:
2.598
Mongi Moujahed1, Mondher Hajji1, Hechmi Benazza1, Mohamed Jemli1 and Mohamed Boussak2
1
Department of Electrical Engineering, Ensit, University of Tunis, Tunis, Tunisia and 2Department of Electrical
Engineering, ECM, University of Marseille, Marseille, France
Anti-Cancer Agents in Medicinal Chemistry
Abstract: Background: In this study, a new model-based fault detection and isolation (FDI) strategy of
IGBTs open circuit faults detection and diagnosis in voltage source inverter (VSI) is proposed for space
vector modulation direct torque control (SVM DTC) permanent magnet synchronous motors drives.
ARTICLEHISTORY
Received: May 9, 2017
Methods: The method is based on the bridge leg pole voltage, when with an additional simple circuit,
Revised: September 17, 2017 the diagnostic method detected and effectively identified single and multiple open-circuit faults of
Accepted: September 17, 2017
inverter IGBTs. Also, a fault compensation strategy is suggested by using the information of the model-
DOI: based FDI stage composed of a four-leg inverter in which the fourth inverter leg is used as hardware
10.2174/2352096510666171108145717
redundancy.
Results: The simulation results of fault detection, identification and compensation methods are included
to prove that the proposal leads to good results.
Keywords: Permanent magnet synchronous motor (PMSM), Fault tolerant control (FTC), fault inverter, fault detection
isolation (FDI), SVM-DTC.
1. INTRODUCTION In this phase, the drive removes the faulty device and
reorganizes the system back to its safety mode and continues
Decades ago, and now scientist lend great attention to its performance [11]. It is worth noting that the first two
fault diagnosis of voltage source inverters (VSI) because tasks (fault detection and fault identification) are primordial
they are essential parts in variable speed drives (VSD) that and called fault diagnosis. We suggest a strategy of fault
are functional in lots of industrial systems [1]. Thanks to compensation by using the data of the model-based stage.
their applications, permanent magnet synchronous motors This is shown in [12-14]; it is formed by a four-leg inverter
yield a set of advantages such as high power density, torque where the fourth inverter is hardware redundancy.
to inertia ratio, efficiency, simple control and cheap
maintenance [2-5]. Also VSD comes first when it is a matter The results in simulation prove that even after the fault
of safety to human life in industrial drives-controlled the system works continuously and with stability, thanks to
environment as a small fault or accident could lead to serious the suggested algorithm.
risks for the safety of high performance systems [6, 7]. The
role of the drives is to stop in emergency any system when in 2. MATERIALS AND METHODS
risk or for maintenance reasons. Besides, the need to develop
reliable drives help reduce the costly unexpected 2.1. VSI Open-Circuit Faults Analysis
maintenance of some machines [8, 9]. For these reasons, a
control system called “fault tolerant control system” FTC is
developed to minimize faults to zero [10]. It takes in charge T1 T
T3
three essential tasks:-detecting faults, identifying then and Vdc/2 D1 D3T5 D5
taking remedial actions. ia ib ic
0 A B C
- fault detection consists in checking whether the system is Th
T2 T4 T6
healthy or not. Next comes the fault identification which Vdc/2 D2 D4 D6
is executed to identify the fault location, its type and
nature. The final phase is the remedial actions or the fault M
isolation.
*Address correspondence to this author at the Department of electrical
engineering, Ensit, University of Tunis, P.O. Box: 1100, Tunis, Tunisia; Fig. (1). VSI standard topology.
Tel/Fax: +21698474717, +21677366486; Fig. 1 shows the common structure of a VSI. The power
E-mail: [email protected]
switches are obtained by the use of IGBTs (T1~T6) with
antiparallel diodes (D1~D6). T1 is an open-circuit failure T2 open circuit fault occurs in two cases when using the
with the antiparallel diode D1 still conducting. The switches status and current loop to detect low IGBT.
diagnostic model uses bridge leg voltage and switching Case1:ia<0, ib<0, T2=1, T4=1, T6=1.
signals based on an analytical model of the VSI employ
bridge leg voltages and switching signals based on an Case2:ia<0, ib<0, T2=1, T4=1, T6=0.
analytical model of the VSI. It is described as follows: uref2 is set between -Vdc and -Vdc/2 (relative to uAM).
This method is formed with the bridge leg pole voltage Consequently, the output signal VJ2 of the no gate varies
(uAM, uBM, uCM). from the low level to the high level. When combining the
uAM: voltage between the poit A and M (Fig. 1). switching signal T2, the open-circuit fault of T2 is obtained
by the Boolean signal as demonstrated in Equation (8).
uBM: voltage between the point B and M
FT 2 T 2VJ 2 (8)
uCM: voltage between the point C and M.
The bridge leg voltage uAM changes when a single but From the previous analysis we conclude that the obtained
various IGBT open circuit faults occur. Consequently, the faults data are based on the switching signals and measure
switches and the current loop as T1 occur. If the phase bridge leg voltage.
currents ia and ib are positive, and the gate signals T1 and T3
To differentiate the bridge leg voltage under faulty
are at a high level; the bridge arm pole voltage u AM is equal
conditions from the normal voltage, an extra simple
to Vdc in normal working conditions. In VSI having a type
hardware circuit is shown in Fig. 2 [15].
star connected load, the three phase currents create a
relationship shown in (1). uAM Sampling VJ1
FT1
+
ia ib ic 0 (1) Link
-
T1
The positive phase A half current is lost when T1 is
uref1
related with an open-circuit failure. So:
ib ic 0 (2) VJ2
FT2
+
in the case 1, where (ia>0, ib>0, T1=1, T3=1, T5=1) -
T2
ia becomes zero and ic is negative according to Equation
uref2
(2). Then ic circulates through D5. Therefore:
u u 0
BO CO (3)
Fig. (2). Proposed diagnostic method.
Then the bridge leg pole voltage in case 1 uAM_case1, can The proposed method seems to be very effective and can
be expressed as: successfully indicating IGBTs faults.
u AM case1 Vdc (4)
3. FAULT COMPENSATION STRATEGY
For case 2, (ia>0, ib>0, T1=1, T3=1, T5=0), ic is negative
and circulates through T6. So, the bridge leg pole voltage in In this part, we suggest a fault compensation strategy by
case 2, uAM_case(2) can be written as: using data of the model-based FDI stage. This is shown in
[16]. It consists of a four-leg inverter and the fourth leg
Vdc
u AM case 2 (5) inverter function as hardware redundancy. Switches ISi and
2 THi are for all i = {1, 2, 3}. The motor phase is linked to the
back-up fourth-leg for its continuous operation. The
Then, the bridge leg pole voltage uAM is be given by: characteristics of greatest importance of this fault
compensation are: First, only one faulty leg can be
Vdc
u V T5
AM dc T5 (6) compensated. Then, there is no need for a control algorithm
2 of reconfiguration. Finally, two back-to-back thyristors are
required (by leg) to replace the faulty-leg.
Where T5 is the switching signal of the IGBT T5, and
is the complementary signal of T5. 4. RESULTS AND DISCUSSION
The value of uref1 is set between Vdc/2 and Vdc (relative to
uAM). A result, the output signal VJ1 of the Not gate varies 4.1. Simulation Results
from the low level to the high level. When combining the T1
switching signal, T1 open-circuit fault is obtained by the Simulations are conducted to prove the feasibility of the
Boolean signal as demonstrated in Equation. (7). fault diagnostic method. The control algorithm set on the
space vector modulation direct torque control (DTCSVM)
FT 1 T 1VJ 1 (7) strategy has been applied by Matlab-simulink use. The
inverter is modeled with a power switches real model and by
Where FT1is the indicator signal for an open-circuit fault pass-diodes. A time of 4 is set on to avoid the short circuit
of T1. in the inverter legs.
Open-Circuit Fault Detection for IGBTs Based on Leg Voltage Recent Advances in Electrical & Electronic Engineering, 2018, Vol. 11, No. 1 23
Table 1 shows the nominal parameters of the permanent Fig. 5 shows the simulation results of the phase currents,
magnet synchronous motors. the leg voltage uAM, and the commanded fault for an open
circuit fault of the upper IGBT T1. In this case, the fault was
T1 T3
T5 T7 applied at t0=1.8s during the positive alternation of the phase
current ia. Its detection is accomplished at t1 = 1.8005s Fig.
0 A B C (6). Therefore, the fault is detected after 0.5ms. Similar results
T2 T4 T6 T8 are obtained when the lower transistor T2 of the first leg is
faulty, as shown in Fig. 7. The fault has been occurred when
M the phase current ia is positive at t0=1.8s. Its detection is
IS1 IS2 IS3 accomplished at t= 1.808s. Note that this delay depends on the
instant in which the fault occurs.
TH
8
Normal operation Fault operation
TH2 6
TH 4
Currents (A)
0
-2
-4
. -6
Fig. (3). Configuration of the VSI post-fault operation.
-8
Table 1. Nominal parameters of PMSM. 1.65 1.7 1.75 1.8 1.85 1.9 1.95 2
Time (s)
Parameters Specification Fig. (5). Stator currents in case of normal and faulty mode.
1.5
Rs 0.5Ω Rated power 1.570kW
Ld 3.9 mH Rated voltage 400V
Lq 3.7 mH Rated current 5.9A 1
Commanded fault
0
Fig. 4 shows the proposed space vector modulation direct
torque control of PMSM drive and open-circuit faults
detection and isolation. -0.5
1.65 1.7 1.75 1.8 1.85 1.9 1.95 2
Vsαref Time (s)
Sa
s-ref
Reference Fault Fig. (6). Commanded fault response.
Space
∆ Voltage Vector Sb Tolerant
Vsβref Modu-
IP PI Inverter
Calculation lation
Φ I uAM
Vs Is F F PMSM
Fig. (8). Stator currents in case of FTC mode. Fig. (11). Wave form of rotor speed in faulty mode.
REFERENCES
[1] M. Boussak, “Implementation and experimental investigation of
sensorless speed control with initial rotor position estimation for
interior permanent magnet synchronous motor drive”, IEEE Trans.
Power Elect., vol. 20, no. 6, pp. 1413- 1422, 2005.
[2] R. Errabelli and P. Mutscher, “Fault tolerant voltage source inverter
permanent magnet drives”, IEEE Trans. Power Elect., vol. 27, no.
2, 2011.
[3] M. Moujahed, H. Benazza, M. Jemli and M. Boussak,” Estimation
by using EKF Techniques for Sensorless DTC of PMSM with load
torque Observer”, Int. Rev. Electr. Eng., vol. 9, no. 2, pp. 270-279,
2014.
[4] M. Caner, C. Gerada and G. Asher, “Permanent magnet motor
design optimisation for sensorless control”, Int. Rev. Electr. Eng.,
vol. 8, no. 1, pp. 172-181, 2013.
[5] M. Moujahed, H.B. Azza, M. Jemli and M. Boussak, “Fault
detection and fault-tolerant control of power converter fed PMSM”,
Electron. Electric. Engineer., vol. 98, no. 2, pp. 121-131, 2016.
[6] G.S. Buja and M.P. Kazmierkowski, “Direct torque control of
PWM inverter-fed AC motors: A survery”, IEEE Trans. Ind.
Fig. (13). Waveform of electromagnetic torque in case of faulty mode. Electron., vol. 51, no. 4, pp.744-757, 2004.
[7] J. Guitard, F. Richard and K. Bouallaga, “Fault tolerant inverter
with real time monitoring for aerospace applications”, IEEE 14th
Internat. Power Electron. Motion Cont. Conf., EPE-PEMC, Ohrid,
Macedonia, 2010.
[8] I. Bahri, I.S. belkhoja and M. Eric, “FPGA based real time
simulation of fault tolerant current controllers for power
electronics,” IEEE Internat. Sympos. Indust. Electron., Seoul,
South Korea, 2009.
[9] M. Trabelsi, M. Boussak and A. Chaari, “High performance single
and multiple faults diagnosis in voltage source inverter fed
induction motor drives,” IEEE XXth Internat. Conf. Electric.
Machines, Marseille, France, 2012.
[10] M. Boussak, “Implementation and experimental investigation of
sensorless speed control with initial rotor position estimation for
interior permanent magnet synchronous motor drive”, IEEE Trans.
Power Elect., vol. 20, no. 6, pp. 1413-1422, 2005.
[11] Y. Wang, L. Zhan, X. Minghui and M. Hao, “A comparative study
of two diagnostic methods based on the switching voltage pattern
for IGBT open-circuit faults in voltage-source inverters”, J. Power
Electronics, vol. 16, no. 3, pp. 1087-1096, 2016.
[12] U.M. Choi, F. Blaabjerg and K.B. Lee, “Study and handling
methods of power IGBT module failures in power electronic
Fig. (14). Waveform of electromagnetic torque in case of FTC mode. converter systems,” IEEE Trans. Power Electron., vol. 30, no. 5,
pp. 2517-2533, 2015.
[13] W. Sleszynski, J. Nieznanski and A. cichowski, “Open transistor
fault diagnosticsin voltage source inverters by analyzing the load
CONSENT FOR PUBLICATION currents”, IEEE Trans. Ind. Electron., vol. 56, no. 11, pp. 4681-
4688, 2009.
Not applicable. [14] D.R. Espinoza-Trejo, D.U. Campos- Delgado, G. Bossio, E.
Barcenas, J.E. Hernandez-Diez and L.F. Lugo-Cordero, “Fault
diagnosis scheme for open-circuit faults in field-oriented control
CONFLICT OF INTEREST induction motor drives”, IEEE IET Power Electron., vol. 6, no. 5,
pp. 869-877, 2013.
The authors declare no conflict of interest, financial or [15] M. Trabelsi, M. Boussak, P. Mestre, M. Gossa, “Pole voltage
otherwise. based-approach for IGBTs open-circuit fault detection and
diagnosis in PWM-VSI-Fed induction motor drives”, IEEE Xplore
Internat. Conf. Power Engineer. Energy Electric. Drives, Malaga,
ACKNOWLEDGEMENTS Spain, 2011.
[16] M. Moujahed, H. Benazza, M. Jemli and M. Boussak,
Declared none. “Implementation of improved sliding mode observer and fault
tolerant control for a PMSM drive”, J. Circuit Syst. Comp., vol. 26,
no. 02, pp. 1750032 , 2017.