On-Line Monitoring and Classification of Stator windings ...journal.esrgroups.org/jes/icraes/CDICRAESFinal/ICRAES16ProcPaper… · On-Line Monitoring and Classification of Stator

On-Line Monitoring and Classification of Stator windings Faults in

Induction Machine Using Fuzzy Logic and ANFIS Approach

Abstract-- the induction machines drives becomes more

and more important used in many industrial applications.

Their attractiveness is largely due to their simplicity,

ruggedness and low cost manufacture, easy maint00enance,

high power efficiency and high reliability, are susceptible to

various types of electrical and/or mechanical faults that can

lead to unexpected motor failure and consequently impulsive

downtime. This made necessary the monitoring function

condition of these machines types for improved an

exploitation of the industrial processes. The aim of this task

is the proposal of a monitoring strategy based on the fuzzy

logic inference system (FIS) and the neuro-fuzzy inference

system (ANFIS) for monitoring and classification of

electrical faults types, especially the open phase and inter-

turns short-circuit in the stator windings. The principle

adopted for the strategy suggested is based on monitoring of

the average root mean square value of stator current (RMS).

Mathematical models and simulations results are presented

to validate the efficiency of this approach.

Index Terms-- Monitoring; Classification; FIS;

ANFIS; RMS.

1. INTRODUCTION

Different of electrical machines types are present in

several processes and industrial equipments. But the

induction machines are currently the principal means in

the industrial sector for conversion electrical energy into

mechanical driving and they are play important roles in

various industrials processing. Though their low cost,

simple maintenance, from the reliability and robustness

perspective point [1, 2].

Although all these advantages, these machines are

easily prone to failure since are frequently installed in

variety and the hostile environment that may be easily led

to the deterioration. Moreover, several problems may

occur during their function because of thermal,

mechanical and electrical stresses, incorrect functioning

condition or manufacturing defects [3].

In recent years the online monitoring and diagnosis

techniques of faults found in three-phase induction

machines are study under various approaches by many

research tasks, since of its considerable interest for the

continuity of the industrial processes service [4, 5].In

specifically the most common electrical faults in induction

machines are related to the stator windings, as inter-turn

shorte circuit account for more than 30% of all faults, also

the open stator phase default is one of faults in stator [7, 8]

Early faults detection allows to minimize the

downtime, the turn-around time of the process in question,

to avoid the damaging consequences, and to reduce the

financial losses [9].

The majority of the monitoring approaches are based on the analysis of electromagnetic magnitude such that the

magnetic flux, the stator or rotor current, and the neutral

voltage [10, 11]. In this case, by measuring accessible and

easily quantifiable magnitudes, includes the stator currents

of the induction machine for calculate their RMS values to

analyze them in a minimum of time and to conclude the

state of the induction machine [12].

However, through this work, we will be interested

particularly in the open circuit and short-circuit inter-turns

faults in stator winding of the induction machine (IM).

The inter-turn short circuit fault in stator windings can

propagate and can be developed either due to total defect insulation inter-turns of stator winding, leading to phase to

ground or phase to phase faults. Some importance is

therefore attached to the early detection of stator faults

[13, 14].

So, the approach that we propose is based on the fuzzy

logic inference system (Fis) and Adaptive Neuro-Fuzzy

System Inference (ANFIS), in order to increase the

efficiency and the reliability of the on-line monitoring and

classification faults in the supervision of the induction

machine [15, 16]. The models of the approach as well as

the global model are simulated by using software

MATLAB®/SIMULINK and the obtained results of

simulations in a healthy function and short-circuit or open

phase faults are presented and interpreted.

Merabet. Hichem

Research Center in Industrial Technologies (CRTI) P.O.

Box 64, Cheraga, Algeria.

[email protected]

Bahi. Tahar

Electrical Department, University of Annaba, Algeria

[email protected]

Drici. Djalel

Research Center in Industrial Technologies (CRTI) P.O.

Box 64, Cheraga, Algeria.

[email protected]

Bedoud. Khouloud

Research Center in Industrial Technologies (CRTI)

P.O. Box 64, Cheraga, Algeria.

[email protected]

Boudiaf. Adel

Research Center in Industrial Technologies (CRTI) P.O. Box 64, Cheraga, Algeria.

[email protected]

Proceedings of the International Conference on Recent Advances in Electrical Systems, Tunisia, 2016

ISBN: 978-9938-14-953-1 (187) Editors: Tarek Bouktir & Rafik Neji

2. MODEL OF HEALTHY INDUCTION MACHINE

The stator and rotor voltage equations can be expressed

in the form following [17]:

!"#$% = &%'"#$% + (() *"#$%,,,,,,,,0 = &'"#$- + (()*"#$- . (1)

The stator and rotor flux equations can be expressed in

the form of following:

/*"#$% = 1"#$% 2 '"#$% + 1"#$%-2 '"#$-,*"#$- = 1"#$-%2'"#$% + 1"#$- 2 '"#$- . (2)

The electromagnetic Torque equation:

345 = 67 2 89:;< 9>;<?2 @ [1%%] [1%-][1%%]A [1--]B 2 C 9:<9><D (3)

The differente inductances matrixes are obtained by:

1%% =EFFFG1H%I1%5 J 67 1%5 J 67 1%5J 67 1%5 1H%I1%5 J 67 1%5J 67 1%5 J 67 1%5 1H%I1%5KLL

LM (4)

1%- =

EFFFG 15 cos N- 15 sin ON- + 7PQ R 15 cos ON- J 7PQ R15 cosON- J 7PQ R 15 cosN- 15 cos ON- + 7PQ R15 sin ON- + 7PQ R 15 cos ON- J 7PQ R 15 cosN- KL

LLM,,,,STU

1-% =

EFFFG 15 cos N- 15 sin ON- J 7PQ R 15 cos ON- + 7PQ R15 cosON- + 7PQ R 15 cosN- 15 cos ON- J 7PQ R15 sin ON- J 7PQ R 15 cos ON- + 7PQ R 15 cosN- KL

LLM,,,,SVU,,,

1-- =EFFFG1H-I1-5 J 67 1-5 J 67 1-5J 67 1-5 1H%-I1-5 J 67 1-5J 67 1-5 J 67 1-5 1H-I1-5KLL

LM (7)

With;

Lsm: mutual inductances of the stator tow-phase winding;

Lrm: mutual inductance of the rotor tow-phase winding;

Lm: the self mutual inductance of the stator and rotor.

3. SHORT-CIRCUIT MODEL OF INDUCTION MOTOR

The models of induction machine under stator inter turn

short circuit fault can be expressed as [18]:

!"!#$%&'( = )(*%&'( +

,,- .%&'(

0 = )/*%&'/ + ,,-.%&'/

101111 = )''*'' 111+1 ,,- .''21 (8)

The resistance of the short-circuit winding is determined

by the relationship below:

)'' = 344566 7 )( 1111 (9)

The flux stator winding and short-circuit winding

equations in dq frame:

!"!#

,89:,- = $;( <)>?;( +@.,( + A

BC)>?D cosE,8F:,- = $,( < )>?,( +@.;( + A

BC)>?D sin E,8G:H,- = )D?D < C)>I?,( cosE + ?;( sinE < ?DJ

2 (10)

The stator and winding currents in dq frame

!"!#?(; = .;(K5 <.;/KL + IM/KN + MOKPJ?D cos E?(, = .,(K5 < .,/KL + IM/KN + MOKPJ?D sinE?D 1= I<.%(L + IKQ?(; + KR?/;JcosE +111111111111IKQ?(; + KR?/;Jsin EJSKP

2 (11)

With;

If : Current of short-circuit.

The constant coefficients KTare shown in Table1.

K6 K5 KL KNM(M/ < MOL U

K6M(MOK5M/

LNCM(K6KP KQ KR

LNCMOK6 CM( CMO

Figure1 present the schematic of induction machine

under inter turns short-circuit fault in phase “a”.

4. OPEN-CIRCUIT MODEL OF INDUCTION MOTOR

The following figure shown the case of open stator fault

in phase "a", the stator and rotor voltages of this machine

are unbalanced [19].

Fig 2. Open stator phase fault

Fig.1 Stator windings with short circuit fault

AC

Source

Ias

Ibs

Ics

σs2

σs1Ias

IcsIbs

Indiction Machine



The voltage and flux systems are expressed by equations

(12) and (13) respectively:

!"#$% = &%'()* _+, . -"#$% / 001 . 2"#$%!"#$3 = &3 . -"#$_3 / 001 . 2"#$4*********** 5 (12)

62"#$% = 7%%'()* _+,. -"#$_% / 7%3'()* _+,. -"#$32"#$3 = 733 . -"#$3 / 73%'()* _+, . -"#$% 5 (13)

The resistance and inductance stator are given by the

following matrix:

&%'()* _+, = 89 9 99 :% 99 9 :%; (14)

7%%'()* _+, = <9 >?@7%A >?@7%A9 7B% / 7%A >?@7%A9 >?@7%A 7B% / 7%AC (15)

The mutual inductance matrix stator-rotor and rotor-stator

in faults case are given by:

7%3'()* _+, =7%A < 9 9 9sinDE3 / @FG H sinIE3J sinDE3 > @FG HsinDE3 > @FG H sinDE3 / @FG H sinIE3J C (16)

73%'()* _+, = 7%A K9 sinDE3 / @FG H sinDE3 > @FG H9 sinIE3J sinDE3 / >@FG H9 sinDE3 > @FG H sinIE3J L (17)

With;

**&%'()* _+, : stator resistance in fault case;7%%'()* _+, : stator inductance in fault case;7%3'()* _+,: mutual inductance stator-rotor in fault case;73%'()* _+,: mutual inductance rotor-stator in fault case.

5. MONITORING OF THE STATOR BY FUZZY LOGIC

1. Monitoring system

In this approach, we have used the fuzzy logic

inference system for monitoring of open phase or inter-

turns short-circuit faults in stator winding of the induction

machine. In this stage, we used linguistic variables and the

membership functions for describe the RMS amplitudes of

stator currents. An interface fuzzy system comprising the

rules and the data bases are established to support the

fuzzy inference system. The state of the machine is

monitoring by using the fuzzy logic system [20, 21].

2. Input-output variables of fuzzy system

The RMS of currents (RMS_Ias, RMS_Ibs and RMS_Ics)

and the state of stator, (CM) are respectively selected as

inputs and output variables of the fuzzy system.

All these variables are defined by using the fuzzy set

theory. Figure 3, shows that CM (Condition Monitoring)

interprets the state of the stator as a linguistic variable,

which could be T (CM) = {Healthy Stator, Damaged,

Seriously Damaged, Open Stator Phase}.Each limit in T

(CM) is characterized by a fuzzy subset. The dialog

system:

· CM {(HS)}: interprets that the stator is healthy;

· CM {(D)}: the stator can by present a minor

short-circuit faults;

· CM {(SD)}: that the critical short-circuit fault;

· CM {(OSP)}: that the stator open phase fault.

The variables of input RMS_Ias, RMS_Ibs and RMS_Ics

are also interpreted as linguistic variables, with, T (Q) =

{Very Small (VS), Small (S), Medium (M), Big (B)} as it

is showing in figure 4.

Fuzzy rules membership functions for the input and the

output. These rules are then defined (shown in Table 2), as

follows:

Table2. Fuzzy rules membership function

RulesIf

RMS_Ias

And

RMS_Ias

And

RMS_Ias

Then CMS

01 VS OPs

02 VS OPs

03 VS OPs

04 S S S Hs

05 M M M Hs

06 S M M Ds

07 S S M Ds

08 M M S Ds

09 M S M Ds

10 S M S Ds

11 M S S Ds

12 B SDS

13 B SDS

14 B SDS

Fig.3. Membership functions for output variables

Fig. 4 Membership functions for input variables



6. ADAPTIVE NEURO-FUZZY INFERENCE

SYSTEM

ANFIS is a hybrid controller structure using fuzzy

logic inference system and the architecture of a neural

network having five-layer feed-forward structure [22, 23].

Thus, the ANFIS offers the advantages of learning

capability of neural networks and inference mechanism of

fuzzy logic. A typical architecture of ANFIS having n

inputs, one output, and m rules is illustrated in Figure. 5.

Here x, y, z and up to n are inputs, f is output, the

cylinders represent fixed node functions and the cubes

represent adaptive node functions. This is a Sugeno type

fuzzy system, where the fuzzy if-Then rules have the

following form:

· Rule1: if x is A1 and y is B1,……………..n is k1 then

f1= (p1x+q1y+r1z+……v1)

· Rule2: if x is A2 and y is B2,……………..n is k2 then

f2 = (p2x+q2y+r2z+……v2).

· Rule m: if x is Am and y is Bm,…………..n is km then

f2= (pmx+qmy+rmz+……vm).

The suggested model for classification system is

developed under Matlab/Simulink. Data base is collected

from on-line is used the RMS of stator current signal.

7. SIMULATIONS AND INTERPRETATIONS

The figures 6.a present the output of the fuzzy value

(the decision). This value is included in the interval CM

= {HS [25 50]}, which corresponds to the limits of the healthy stator case. In addition, in the figure 5.b present

the fuzzy inference diagram of the currents phases and

the decision.

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

20

40

60

80

Time (s)

Fuzzy v

alu

es

HS

-a- Output Fuzzy Values

-b- Fuzzy inference diagram

Fig.6 Characteristics of the IM (healthy case)

-a- Output fuzzy Values


Fig.7 Characteristics of the IM with 5% short-

circuit in phase "a" (Damaged case)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

20

40

60

80

Time (s)

Fuzzy v

alu

es Ds

Fig.5. Typical ANFIS structure



The figures 7.a and 8.a corresponding to tests of

short-circuit currents phases, present increases in

magnitudes proportional to the numbers (proportion) of

the short-circuit stator turns. In addition, these figures

indicate values which correspond to those that indicate

the presence of the fault. Indeed, in the case of 5%

shorted-circuit turns the decision indicates CM = {D [45

50]}. On the other hand, for 10 % of shorted-circuit

stator turns the decision indicates CM = {SD [70

100]}.Therefore, these tests validate that the approach is

reliable and exploitable.In addition Figure9.b indicates the fuzzy output values

of stator decisions state, these values are included in the

universe of discourse with CM interval = {open stator

phase OPS [0 25]} what correspond to the open stator

phase case limits. Figure 9.c shows the diagram of fuzzy

inference in the fault of open stator of phase “a”

The trained and checked ANFIS output for fault type’s

classification are shown in Fig. 10. To validate our

network (shown the Fig.11), a test of recognition is carried

out. The input relationships or dependency for the ANFIS

output are in addition analyzed. These are the unique characteristics of adaptive neuro-fuzzy inference system.

The mapping is optimized by neuro adaptive learning

techniques by fuzzy modeling procedure to learn

information about the data set for monitoring the stat of

induction machine in our case study.

1 2 3 4 5 6 7 8-0.5

0

0.5

1

1.5

2

2.5

3

3.5

Index Data

faults t

ype

Testing-Data

ANFIS Output-Test 35% inter-turns

15% inter-turns

Healthy case

5% inter-turns

Fig.11 Testing Data and Testing Output for the ANFIS (Stator inter-turns short circuit)

Hs

Ops

Ds

SDs

0 50 100 150 200 250 300 350 400-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

Index Data

faults t

ype

Training Data

ANFIS Output

Checking Data

ANFIS Checking Output

Healthy case

5% inter-turns

15% inter-turns

35% inter-turns

Fig.10 Training, Testing and Checking Output for the

ANFIS (Stator inter-turns short circuit)

Hs

Ops

Ds

SDs

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

20

40

60

80

Time (s)

Fuzzy v

alu

es

SD

-a- Output Fuzzy Values


Fig.8 Characteristics of IM with 10 % short-circuit inphase "a" (Seriously Damaged case)

-a- Output fuzzy Values

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 510

20

30

40

50

60

70

80

90

Time (s)

Fuzzy v

alu

es

OPs


Fig.9 Characteristics of IM with open stator phase



8. CONCLUSION

In this paper, we presented the development of a fault

model of the induction machine then the simulations of

stator short-circuit and open stator faults.

In firstly we are represented the mathematical

development then the simulation of healthy function and

faulty induction machine models under

MATLAB/Simulink software.

In the second part of this work we have assembled

feasibility to monitoring and classification of open stator phase and inter-turns short-circuit fault inter turns in

stators windings of the induction machine by supervising

the root mean square value of stator currents (RMS)

magnitudes based on the fuzzy logic inference system

(FIS) and the neuro-fuzzy inference system (ANFIS).

In addition, this approach using the artificial

intelligence easily can by to inform about the induction

machine stat, and to predict the fault severity and classify

of faults types.

References

[1] W. Fontes Gody, I. Nunes de Silva, A. Godtel, R. henrique

Cunha Palacios, “ Evaluation of stator winding faults severity

in in inverter-fed induction motors”,Applied Soft Computing

Vol .32(2015) 420-431.

[2] J. Seshadrinath, B.Singh, and B. Ketan Panigrah,

“Investigation of Vibration Signatures for Multiple Fault

Diagnosis in Variable Frequency Drives Using Complex

Wavelets”, IEEE Transaction on power electronics, vol. 29,

no. 2, pp 936-945. Feb 2014.

[3] F. Babaa, A. Khezzar, M.El Kamel Oumaamar,

“Experimental investitgation and comparative study of

inerturn short-circuits and unbalanced voltage supply in

induction machines”, Front. Energy, Vol .7 issue 3, pp 271–

278. 2013.

[4] S.M. Shashidhara and P.S. Raju, “Stator Winding Fault

Diagnosis of Three Phase Induction Motor by Park’s Vector

Approach”, Internatonal Journal of Advance Research In

Electircal Electronics and Instrumentation Engineering,

Vol.2, pp. 2901-2906, Issue, 7, July. 2013.

[5] H. Eristi, “Fault diagnosis system for series compensated

transmission line based on wavelet transform and ANFIS”,

Measurement, vol 46,p p 393–401. 2013.

[6] K. Moin Siddiqui, K. Sahay, V.K.Giri “Health Monitoring

and Fault Diagnosis in Induction Motor”, International

Journal of Advanced Research in Electrical, Electronics and

Instrumentation Engineering Vol. 3, January 2014.

[7] K. Moin Siddiqui and V.K.Giri. “Broken Rotor Bar Fault

Detection in Induction Motors using Wavelet Transform”, Int.

Conf IEEE,[ICCEET], pp. 1-6, Chennai, India, March, 2012.

[8] M. Akar, Ilyas .C. Ankay, “Broken rotor bar fault detection in

inverter-fed squirrel cage induction motors using stator

current analysis and fuzzy logic”, Turk J Elec Eng & Comp

Sci, Vol. 20, No.1, 2012.

[9] A.Verma, S. Sarangi, M.H. Kolekar “Stator winding fault

prediction of induction motors using multiscale entropy and

grey fuzzy optimization methods” Computers & Electrical

Engineering .June 2014.

[10] Somaya A. M. Shehata, Hamdy S. El-Goharey, Mostafa I.

Marei, Ahmed K. Ibrahim “Detection of Induction Motors

Rotor/Stator Faults Using Electrical Signatures Analysis,”

International Conference on Renewable Energies and Power

Quality (ICREPQ’13)Bilbao (Spain), 20th to 22th March,

2013.

[11] D. K. Chaturvedi, Akash Gautam, Mayank Pratap Singh, Md.

Sharif Iqbal“ On Line Fault Identification of Induction Motor

using Fuzzy System”, TECHNIA – International Journal of

Computing Science and Communication Technologies,

VOL.6 N. 2 ISSN 0974-3375). 01/2014.

[12] A. M. da Silva, Richard J. Povinelli, ,and Nabeel A. O.

Demerdash, Life Fellow “Rotor Bar Fault Monitoring Method

Based on Analysis of Air-Gap Torques of Induction Motors ”

IEEE Trans on industrial informatics, vol. 9, no. 4 , pp 2274-

2283. November 2013.

[13] S. M.Ahmed , Haitham Abu-Rub, Shady S. Refaat, Atif Iqbal,

“Diagnosis of Stator Turn-to-Turn Fault and Stator Voltage

Unbalance Fault Using ANFIS”, International Journal of

Electrical and Computer Engineering (IJECE) Vol.3, No.1, ,

pp. 129~135 ISSN: 2088-8708, February 2012.

[14] Y. D. Sanjeev, K. Srivastava, Chris S. Edrington, David A.

Cartes “Application of artificial intelligence to stator winding

fault diagnosis in Permanent Magnet Synchronous Machines”

Electric Power Systems Research 2013, vol 103,pp 201–213.

[15] A. K. Ibrahim, M.I. Marei, H.S. El- Goharey, S.A.M.

Shehata, “Modeling of Induction Motor Based on Winding

Function Theory to Study Motor under Stator/Rotor Internal

Faults,” 14th

Int. Middle East Power Systems Conference,

(MEPCON’10), Cairo, Egypt, December 19-21, 2010.

[16] V. Prakash Pandey and P. K. Choudhary, “Induction Motor

Condition Monitoring Using Fuzzy Logic”, Advance in

Electronic and Electric Engineering. ISSN 2231-1297,

Volume 3, N° 6 pp. 755-7646, 2013.

[17] Y. Soufi, T. Bahi, H. Merabet and S. Lekhchine, “Short

Circuit between Turns in Stator Winding of Induction

Machine Fault Detection and Diagnosis,” Applied Mechanics

and Materials; Vols. 416-417 (2013) p 565-571. 2013.

[18] Lashkari. N., Poshtan, J., and Azgomi, H.F,’ Simulative and

experimental investigation on stator winding turn and

unbalanced supply voltage fault diagnosis in induction motors

using Artificial Neural Networks’ ISA Transactions, Vol. 59,

pp. 334-342, 2015.

[19] Asgari. S.H., Jannati. M., Idris, N.R.N., “I Modeling of three-

phase induction motor with two stator phases open circuit,”

Energy Conversion, IEEE Conference on, pp. 231 – 236.

[20] K. Bhardwaj, A. Agarawal , “ Fault Diagnosis of Three

Phase Induction Motor using Fuzzy Logic Controller and

Sequence Analyzer”, MIT International Journal of Electrical

and Instrumentation Engineering, Vol. 2, No. 2, pp. (112-118)

Aug. 2012.

[21] MERABET. Hichem, DRICI Djalel, AOUABDI Salim,

“Monitoring of Stator windings Faults in Induction

Machine Using Fuzzy Logic,” journal of Electrical and

Electronics Engineering (JEEE),Vol. 9, No.2, pp.49-54, 2016.

[22] Haitham Abu-Rub, S. M.Ahmed , Shady S. Refaat, Atif Iqbal,

“incipient bearing fault detection for three phase bruchless

DC Motor Drive usin ANFIS”, IEEE, 2011.

[23] S. M.Ahmed , Haitham Abu-Rub, Shady S. Refaat, Atif Iqbal,

“Diagnosis of Stator Turn-to-Turn Fault and Stator Voltage

Unbalance Fault Using ANFIS”, International Journal of

Electrical and Computer Engineering (IJECE) Vol.3, No.1, ,

pp. 129~135 ISSN: 2088-8708, February 2012.



Documents

On-Line Monitoring and Classification of Stator windings ...journal.esrgroups.org/jes/icraes/CDICRAESFinal/ICRAES16ProcPaper… · On-Line Monitoring and Classification of Stator