- Design of Hybrid Controller for Direct Torque Control of Induction Motor Drive
- ijee98
- A Microcomputer-based Induction Motor Drive_1999
- 01157850.pdf
- Bai 2013
- Rotor Ident 11
- Final 1
- Inverse Fuzzy
- Paper
- A Robust RF-MRAS Based Speed Estimator Using Neural Network as a Reference Model for Sensor-Less Vector Controlled IM Drives
- Document
- Editing-Sensorless-Flux-Vector-Control-of-Induction-Motor.docx
- DYNAMIC MODEL OF INDUCTION MOTORS FOR VECTOR CONTROL
- 16_msivasathyanarayana_finalpaper--IISTE research paper
- 01375643
- 1639014
- fulltext_140
- Motor and Starter Training
- WEG INVERTER MOTOR _ANTECH
- Yaskawa A1000 Catalog.pdf
- Motor Control
- New Microsoft Office Word Document (2)
- Main Paper
- LCR meter
- Unit 3.pdf
- Induction Motor
- SE Electrical Syllabus
- Servo and Robot Drives
- Identification
- SrikantaMurthy Umn 0130M 17368
- 4_types of dc chopper.pdf
- MIT9_00SCF11_text.pdf
- 1_Architecture of 8085.pdf
- co design.ppt
- CHAPTER 5.pptx
- PSY_chapter__1.pdf
- 77442_92066v00_motor-control-with-arduino-a-case-study-in-design.pdf
- Chapter 4 bps.pptx
- Dream-Psychology-Sigmund-FreudPsychoanalysis-for-Beginners.pdf
- 3_three phase converter.pdf
- Chapter3PPT.pptx
- chap4.pptx
- using-arduino-with-matlab-and-simulink.pdf
- 5_AC motor starter.pdf
- 2_AC-DC Converter.pdf
- Grid.ppt
- anfis
- ELK-1303-143_manuscript_1.pdf
- MECH II TO VIII
- ELK-1212-31_manuscript_1.pdf
- ELK-1303-143_manuscript_1.pdf
- ELK-1305-114_manuscript_2.pdf
- Ita
- ELK-1212-31_manuscript_1.pdf
- Cognitive Psychology.pdf
- ELK-1212-31_manuscript_1.pdf
- pwm
- ELK-1305-114_manuscript_2.pdf
- ELK-1108-38_manuscript_2.pdf

based on GSA for tuning of PI controllers

1

2

**M. Jannati1, A. Monadi1, N. R. N. Idris1, A. A. M. Faudzi2* and F. Blaabjerg3
**

1

UTM-PROTON Future Drive Laboratory, Faculty of Electrical Engineering,

Universiti Teknologi Malaysia, Johor Bahru, MALAYSIA

2

Centre for Artificial Intelligence and Robotics (CAIRO), Universiti Teknologi

Malaysia, Johor Bahru, MALAYSIA

3

Department of Energy Technology, Aalborg University, DK-9220 Aalborg East,

Denmark

3

Abstract:

4

5

6

7

8

9

10

11

12

**This paper discusses d-q and Winding Function Method (WFM) for modeling and
**

Rotor Field-Oriented Control (RFOC) system for controlling faulty three-phase

Induction Motor (three-phase IM when one of the phases is disconnected). In the

adapted scheme for controlling faulty IM, it is necessary for the PI controller

coefficients to change. For this purpose, Gravitational Search Algorithm (GSA) is used

for tuning of PI controllers. Results display the strength and ability of the technique to

improve the performance of the faulty IM control.

13

14

15

16

17

18

19

**Keywords: Faulty three-phase Induction Motor, Gravitational Search Algorithm,
**

Modeling, Rotor Field-Oriented Control, Tuning of PI controllers

20

21

1- Introduction

Variable speed drives can provide reliable dynamic systems and important savings in

energy custom and costs of the Induction Motors (IMs). Developing high performance

control methods for IM drive systems need an accurate motor model. Recently, various

techniques of modeling induction machines have been presented [1-9]. In [1,2] dynamic

mesh reluctances, in [3,4] d-q model, in [5,6] enhanced equivalent circuit and in [7,8]

Winding Function Method (WFM) has been reported for modeling induction machines.

One of the most commonly models which is used for IMs is the d-q model [9],

presented by Park. This method is based on the assumption that the stator windings are

sinusoidal distributed. This supposition caused the harmonics of the windings

distribution to be removed in the machine analysis. A method based on the real

distribution of stator winding for modeling and calculation of mutual inductances,

which is called the WFM, was proposed by Toliyat and Lipo [7,8]. One of the

advantages of WFM is the ability to model the fault conditions such as cut-out fault in

IMs. In [10-14], the WFM is used to analyze some common faults in IM such as broken

*

**Corresponding Author: A. A. M. Faudzi
**

athif@fke.utm.my Tel.: +607-5535291 Fax: +607-5566272

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

**rotor bars, cracked rotor end rings, shorting, opening, and abnormal of the stator
**

winding circuit. Several methods have been also presented to develop detection of

faults in IM such as stator winding fault [15,16] and rotor bar fault [17,18]. One of the

major failures in IM is open-phase fault. Opening-phase fault is occurred by the

opening of windings, blown fuse and etc. Based on IM equations, in [19], with

comparison between current space phasor and measured value, In [20], Extended

Kalman Filter (EKF) and unknown input observer, In [21], neural networks and look

up table, in [22], negative sequence current estimation have been shown for detection of

the windings fault of stator in the motor. In [23], a small amplitude three-phase high

frequency signal has been employed for detection of fault in stator windings for

permanent magnet synchronous motor. This method gave fast detection of open-phase

stator winding and is supposed in this work. Field-Oriented Control (FOC) is one of the

best techniques for controlling the torque and speed in IMs. This method separates

motor current into torque and flux producing components. The torque of motor is

proportional to the product of these two perpendicular components and they can be

treated independently. This means that by using this method, the IM control is

transformed to an easy control system similar to the DC motors [24]. Faulty IM control

is clearly different from the balanced three-phase IM control. By using the usual

balanced IM control, for faulty IM, considerable ripples in the torque and speed

response will be present [3,4]. In 1996, Zhao and Lipo presented a method for modeling

and controlling of a multi-phase IM based on FOC when one of the stator phases is cut

off [25,26]. In addition to that in [27] presented an approach for modeling (d-q model)

and controlling (FOC method) of a dual three-phase IM when two stator phases are cut

off. In [3,4], Jannati et al. showed how conventional FOC technique can be used for

faulty three-phase IM control. The authors showed that with some changes in the

conventional vector control, unbalanced three-phase IM control is possible. In the

process of obtaining the FOC equations however, the backward components of the

stator voltage equations have been neglected [3,4]. In this paper, d-q and WFM are

discussed for modeling of faulty IM (three-phase IM under open-phase fault).

Moreover, a new method for vector control of faulty three-phase IM based on the Rotor

Flux Oriented-Control (RFOC) method is checked by simulations. In the proposed

controlling method, two new rotational transformations are used to the equations of

faulty IM. By using these matrixes, the faulty motor equations become similar to the

balanced equations. The research potential of the presented drive system for faulty

motor is particularly towards the development of PI controllers. Therefore, optimization

of the PI controllers for having high performance of the presented controlling method is

essential. Some important optimization algorithms techniques are as follows:

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

**Central Force Optimization (CFO) [28], Genetic Algorithm (GA) [29], Artificial
**

Immune Systems (AIS) [30], Particle Swarm Optimization (PSO) [31,32], Simulated

Annealing (SA) [33], Ant Colony Optimization (ACO) [34], Bacterial Foraging

Algorithm (BFA) [35,36] and Gravitational Search Algorithm (GSA) [37], are

evolutionary algorithms which have been proposed for global optimization e.g.

complex nonlinear problems. These methods are increasingly analyzed or improved by

74

75

76

77

78

79

Besides. Rotor and stator flux winding's flux axes under open-phase fault can be presented as Fig. d-q model of faulty three-phase IM with unequal stator windings is presented. the main idea of using rotational transformations for faulty three-phase IM control and RFOC equations for faulty motor by using novel transformations is expounded in this part. PSO and CFO. A brief overview of the GSA algorithm is presented in part 5. It models the masses that attract each others in search space based on gravity lows which means that heavy mass has greater gravitational force and move slowly. Each mass is a problem solution so heavier mass is related to good solution. starting from some solution points called population and determination of a global solution for problem based on the nature's rules. Based on Fig. These algorithms are inspired by nature. The simulation results are shown in part 6 and part 7 concludes the paper. Some efforts have been made to compare the performance of these algorithms with each other [37. PI controller gains are obtained off-line by GSA based on the speed error. θr is the angle between “as” and “ar” axes and θo is the angle between “as” and “ds” axes. 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 2. 43-47]. GSA is based on the Newtonian laws of gravity and was first introduced by Esmat Rashedi et al [37]. After that. The proposed method in this paper not only can be used for critical industrial applications where we need to have a faulttolerant drive system but also can be employed for vector control of single-phase IM (the single-phase IM can be considered as an unbalanced two-phase IM).d-q Model for Faulty Three-Phase IM The faulty IM drive can be shown as Fig. 106 107 In Fig.researchers in different areas [38-42]. 2. the vector control equations for faulty IM by using conventional transformations are presented in part 4. 2. The presented method in [37] has been compared with some familiar heuristic search approaches such as GA. 104 105 Sinusoidal waveform is considered for spatial distribution of the windings. 108 109 o r (1) According to Fig. This paper is organized as: In part 2. 1 (assume that a phase cut out happened in phase “cs”). the normalized transformation for the stator and rotor components is obtained as [4]: TS 1 / Tr 1 1 2 1 1 cos cos 2 / 3 cos 4 / 3 2 / 3 sin sin 2 / 3 sin 4 / 3 110 111 112 (2) 113 (3) 114 . In this research. 2. 2. The WFM model of faulty three-phase IM is presented in part 3. In [37] the authors showed the GSA is more accurate and lead to greater benefits in finding a global minimum point.

λsds and λsqs are the stator d-q axes fluxes and λsdr and λsqr are the rotor d-q axes fluxes.5Lms 134 M q 3 / 2 Lms . the difference between d-q model of the balanced and faulty three-phase IM equations are summarized in Table (1). isds and isqs are the stator d-q axes currents isdr and isqr are the rotor d-q axes currents. τl. sqs Lqs i qss M q i qrs (6) 119 sdr M d i dss Lr i drs . In the analysis using WFM. Actually.WFM Model for Faulty Three-Phase IM An alternative method that can be used to model IMs is by using the WFM. M d 1. Lds = Lqs = Ls and Md = Mq = M in the equations of the faulty motor. rqs and rr indicate the stator and rotor resistances. Lqs. we can obtain the familiar equations of balanced motor. J and F are electromagnetic torque. 3. d-q model for faulty motor in the stator reference frame (superscript “s”) are obtained as follows (see Appendix A): d dsds s . Lr Llr 1.5Lms In summary. rds. For three-phase IM with balanced stator windings in the faulty mode. the equations of the three-phase IM under open-phase fault have the same structure faulty with the equations of the balanced three-phase IM. Lds.5Lms Lqs Lls 0. Md. As can be seen from equations (4-9). Lds Lls 1. load torque. Mq and Lr indicate the stator and rotor self and mutual inductances. the actual spatial distribution of windings is considered. by substituting rds = rqs = rs. vsds and vsqs are the d-q axes voltages. In this method. τe. we have [4]: 123 124 125 126 127 128 129 130 131 132 133 rs rds rqs .5Lms . vqs rqs iqss qs dt dt 115 116 s vdss rds idss (4) 117 dsqr dsdr s s 0 ri r qr . the following assumptions are normally applied: 135 136 137 138 139 140 Uniform air gap 141 Same rotor bars 142 Skew effect is neglected 143 .Applying the transformation vectors ([TS] and [Tr]). inertia and viscous friction coefficient and r is the motor speed. sqr M q i qss Lr i qrs (7) 120 Pole ( M q i qss i drs M d i dss i qrs ) 2 (8) 121 (9) 122 s r dr e d Pole e l J r F r 2 dt where. 0 rr iqr r sdr dt dt (5) 118 sds Lds i dss M d i drs .

Lss. Ir 0 T i I s i1s 149 T 0 0 T Rs. “nB” is the turn function of the winding “B” and “NA” is the winding function of the winding “A”. 5 and equation (11). s Lss I s Lsr I r . Lsr and Lrr can be calculated for balanced three-phase IM. 3. T sr s d r dt Lsr Te I Ir r (10) 145 146 147 T s . z. z. z. From Fig. 144 Saturation is negligible The equations of balanced three-phase IM with “m” rotor bars can be written as follow [7. 6 respectively. 4 show the turn function for the stator phases and the first rotor bar for the simulated balanced motor respectively (for the second rotor bar. Lrr and Lsr are defined in Appendix A. the waveform of Fig. where. 3 and Fig. n(φ) is the turn function and ˂ n(φ) ˃ is the average value of turn function. 4. 3 and Fig. The mutual inductance between windings “B” and “A” (LBA) in terms of turn function and winding function is calculated by following equation [7]: 2 LBA rm o r 0 n . Fig. 4.8]: d s dt d Vr Rr I r r dt Vs Rs I s . “l” is the length of tack. T vmr T i2s i3s r 1 i2r imr . Equation (10) can be written as: s v1s v2s v12 rs s s s v2 v3 v23 0 rs rs i1s 0 s d i2 rs s dt i3 s1 s2 s s 2 3 (12) 165 166 where: s s s1 s2 L11 Ls21 L12 Ls22 s s s s s s 2 3 L21 L31 L22 L32 0 1 1 161 162 163 164 s sr L13 Ls23 i1s L11 Lsr21 Ls23 Ls33 i2s Lsr21 Lsr31 1 i3s 0 ir sr sr L12 Lsr22 L128 Lsr228 1r i Lsr22 Lsr32 Lsr228 Lsr328 2 0 0 ir 28 167 .< n(φ)>. Therefore. the winding function is defined as: N(φ) = n(φ) . based on Fig. 2. dzd l 1 B rm A rm rm (11) 150 151 152 153 154 155 156 157 158 159 160 0 where “r” is the average radius of the air gap. The motor that is investigated in this paper has 28 rotor bars and 36 stator slots. “g” is the air gap function. g . Rr. 5 and Fig. N . 148 where: V v Vs v1s s 1 s r 1 r v2s s2 v1r T v3s s3 . 4 is shifted to the right by 2π / 28 = π / 14). In WFM. the winding function of the stator phases and the first rotor bar are shown in Fig. r Lrr I r L I r . Lss.

the following equations are obtained: 175 176 177 Stator d-axis voltage equation: 178 (14) 179 180 Stator q-axis voltage equation: (15) 181 182 Rotor d-axis voltage equation: (16) 183 . The only difference is that.Equations of faulty motor have the same formation to the balanced motor equations. in the faulty mode.RFOC for Faulty Three-Phase IM In the RFOC method. 171 4. by using following matrix. θe is the angle between the stationary reference frame and rotating reference frame By using conventional transformation to the faulty motor (equations (4-9)). machine equations are transferred to the rotating reference frame [24]: T cos sin e s e e sin e cos e 168 169 170 (13) 172 173 174 where. the row and column for the phase “c s” are eliminated.

qr sin e sin e idss . there are not the backward terms).qr idse. jIf. If and jVf. By applying equation (18). In this case. All the parameters in Fig. rotational transformations for the stator variables that resolve this problem are proposed.dr cos e sin e idss . Vb Vm j a 2 a 2 a 1 1 I f I m jaI a . The backward components are generated because of different inductances in the faulty IM model equations (Md ≠ Mq and Lds ≠ Lqs). 8 are defined in [48].qr 187 In general.dr cos e iqss . The faulty IM control can be done by controlling forward and backward components independently but the control scheme will be very complex. Vf are the currents and voltages of the balanced IM. 8 [48].dr e s iqs . this figure can be changed into a balanced circuit. equations (14-17) include two terms. 9. 9. 9. a simplified circuit can be shown as Fig.dr cos e e iqs . “+e”: forward components and “-e”: backward components as shown in Fig. In this paper. I b I m jaI a 2 2 (18) 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 From Fig.qr sin e cos e iqs . which are displaced from each other by 90. with neglecting of Zd. if we ignore Zd. The main idea of this transformation is obtained from the steady-state equivalent circuit of the single-phase IM. Equation (18). V V 1 1 V f Vm j a . The equivalent circuit of single-phase IM is shown in Fig. This motor is generally unbalanced with two unequal main and auxiliary windings. can be re-written as: 204 205 206 207 .184 Rotor q-axis voltage equation: (17) 185 186 where: ids e. 1Z Z d la2 Z lm 2 a (19) Based on Fig. Each of these terms (forward or backward terms) is like a balanced motor which turns in the forward or backward direction (in the equations of balanced IM. 7.

Lds = Lls + 3/2 Lms. By considering Lqs / Lds = (Mq / Md)2 (in d-q model of faulty IM: Md = 3/2 Lms. rotor voltages. flux and torque equations become like balanced IM equations (the only difference is that in the balanced mode we have : M = 3/2 Lms and in the faulty mode we have Mq = √3/2 Lms). Va vds . From equation (20): 209 210 Proposed rotational transformation for stator voltages: 211 e T vs e ds e qs s ds s qs Lqs cos e Lds Lqs Lds sin e sin e s ds qss cos e (21) 213 Proposed rotational transformation for stator currents: i i e ds e qs i e Tis i s ds s qs Lds cos e Lqs Lds sin e Lqs sin e s ids iqss cos e (22) N 1 Nm q a Na Nd 214 215 In our deduction following substitutions is used: 1 cos e . rotor voltage and stator voltage without supposition of Lqs / Lds = (Mq / Md)2 are given in Appendix B). RFOC equations for faulty IM are gained as: (24) 229 230 . f j 1 Vm I b ja 1 I m a (20) 208 Equation (20) indicates transformation matrixes for changing the variables from unbalanced mode to balanced mode (e. Mq = √3/2 Lms. Lqs = Lls + 1/2 Lms and Lms ˃˃ Lls ) which is equivalent to neglecting inductance of Zd in Fig. by simplifying the single-phase IM equivalent circuit. jI f ids . 217 218 219 220 221 222 223 224 225 226 227 In the RFOC method [24]: 228 edr r . Va and Vm to Vf and jVf). 212 Lqs Lds e e s s V f vqs . the equations of stator voltage include forward terms and backward terms (equations of flux. eqr 0 Consequently. jV f vds . I m iqs As it is shown. 1 j V jI a jV f a j Ia a f V 1 . 9. Vm vqs (23) 216 e e s s I f iqs .g. However. two novel rotational transformations are obtained. It is expected by using of these matrixes (equations (21. equations of faulty IM became like to balanced equations.. j sin e . I a ids .22)) to the faulty IM equations. torque.

Tr (Tr = Lr / rr) is the rotor time constant. we have: rs. the PI controllers should be changed from the balanced mode to the unbalanced 247 248 249 . 234 235 236 237 238 239 240 241 242 243 244 In conclusion. Based on equations (25) and Table (2). the red colors illustrate the sections of the conventional controller which require to be changed for faulty motor control. the only difference between equations of conventional RFOC and modified RFOC is that in the modified RFOC equations it is obtained: rs = ((Mq2 + Md2) / 2 Md2) rs) = 2 / 3 rs. vds-e and vqse as above but in the conventional RFOC equations. M = 3 / 2 Lms. 10. Fig.GSA Algorithm It can be observed that in the modified block diagram for RFOC of faulty three-phase IM. The comparison between stator voltage equations of the conventional RFOC and modified RFOC is summarized in Table (2). In Fig. Moreover. Ls = Lqs = Lls + 1 / 2 Lms. Therefore. vdsd and vqsd is generated using Decoupling Circuit and vdsref and vqsref is generated using PI controller in the vector control block diagram of the faulty IM. the comparison between two vector control techniques is summarized in Table (3). 245 246 5. Ls = Lls + 3 / 2 Lms).M q idse r 1 Tr d / dt e r M q iqse Tr r pole M q r iqse 2 Lr e (25) 231 e d ref e vds vds vds vds e d ref e vqs vqs vqs vqs 232 where: d vds eiqse ( Lqs ref vds ( ref vqs ( Lr rds M q2 rqs M d2 2M d2 v i ( Lqs d qs M q2 e e ds M q2 Lr M q M qidse r )( ) Lr Tr )idse ( Lqs ) e M q rds M q2 rqs M d2 2M d2 )( M q2 didse ) Lr dt r 233 Lr )iqse ( Lqs e vds rds M q2 rqs M d2 cos 2 e ) e ( 2M d2 vqs sin 2 e M q2 diqse ) Lr dt sin 2 e idse cos 2 e iqse In equation (25). 10 can be recommended for RFOC of faulty IM. M = Mq = √3 / 2 Lms.

Go and α are user defined constant and T is the total number of iterations and t is current iteration.2.…. xin for i=1. The outcomes of the total force on agent “i” in dimension d.….. the velocity and position of each agent is updated as follows: 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 vid (t 1) randi vid (t ) id (t ) (30) 266 xid (t 1) xid (t ) vid (t 1) (31) 267 268 The masses of the agents are updated by equations (32-34). 1].. M i M pi M ii M i mi (t ) i=1.N } (35) 271 272 273 274 . randj is a random number in the interval [0. GSA is employed for optimization of the PI controllers. j i j d ij (29) (t ) where.. worst(t) and best(t) are worst and best solution in iteration t and can be defined for minimization as follows: best (t ) min fit j (t ) j{1. The GSA algorithm starts with position of masses (agents) given by equation (26): X i xi1 .2. After each iteration. ε is a small constant and Rij(t) is the Euclidian distance between “i” and “j” agents. In this paper. N fit i (t ) worst(t ) best (t ) worst(t ) M i (t ) mi (t ) j1 m j (t ) N (32) 269 (33) 270 (34) fiti(t) represents the fitness value of the agent “i” at iteration t......mode. G(t ) Go e t T (28) In equation (28). Ma and Mp is the active and passive gravitational mass related to agents “i” and “j” respectively.... xid is the position of ith mass in the dth dimension. N (26) where. is equal to: Fi d (t ) N rand F j 1. At the specific time t the force between mass “i” and mass “j” is defined as equation (27): Fijd (t ) G(t ) M pi (t ) M j (t ) ( Rij (t ) )( x dj (t ) xid (t )) (27) where. G(t) is a gravitational constant which is updated by equation (28). xid ..

by using conventional controller. the proposed controller produces fewer ripples in the speed and torque with superior acceleration (in this case. a step change in the load torque is then introduced (based on [49]. 12 it can be seen that with the conventional RFOC. with the proposed RFOC it only took about 0. The performance of the proposed faulty IM drive system varies according to the PI controller coefficients and it is indicated by the value of the Integral Time Absolute Error (ITAE).5 s.. Then a fault (cut-off in phase “cs”) is employed at t = 0. At the same time. then a load torque equal to 1 N. 10 at the instance the fault is introduced. a vector control drive system according to Fig. The simulated motor is fed by a PWM 3-leg VSI.. the motor is started under healthy condition.Simulation Evaluations 288 To confirm the usefulness of the presented drive system for a faulty IM. 14 (c). whereas in Fig. When t = 4 s. after the fault is introduced. From Fig. The ITAE is considered as an objective function. 12 and 14 shows the simulation results of the conventional RFOC that is utilized even after the faulty condition is introduced.N } During each iteration. the speed oscillation after applying load torque and at steady state is ~ 12 rpm at rotor speed of 500 rpm but by using proposed modified controller 307 308 309 310 311 312 . the conventional RFOC is also simulated. GSA reduces the number of agents based on fitness evaluation and finally one agent remains as the best solution of problem. as can be seen in Fig. WFM is considered for modeling of the unbalanced motor. it is supposed that the fault is happened before the motor is run (like single-phase IM). 13 and 15 shows the simulation results where the control is swap to the one as shown in Fig. However. However Fig. In Fig.m at t = 2. important decrease in the torque and speed ripples can be views in Fig. 14 and 15. The ratings and parameters of the motor used in the vector control are given in Appendix C. the maximum permissible torque in three-phase IM under open-phase fault. The purpose of GSA is to minimize the objective function or maximize the fitness function (fitness function is 1 / (ITAE+1)). the time for the speed to recover to its reference value after a fault is introduced is about 2 s. 289 290 291 292 293 Fig.worst(t ) max fit j (t ) (36) 275 j{1.1 s. 10 is simulated using the MATLAB. However. is about 30% of the rated motor torque). 13. 12 and 13. Results demonstrate that the conventional RFOC method (Fig.5 s is applied to the machine. The principle of GSA is presented in Fig. 11 [37]. 294 295 296 297 298 299 300 301 302 303 304 305 306 In Fig.. 13.. 12) is not capable to control the faulty IM accurately. The parameter settings for the GSA are considered as: 276 277 278 279 280 281 282 283 Population size or number of agents: 50 284 Maximum iteration: 50 285 In equation (27): ε = 0 286 In equation (28): α = 20 & Go = 100 & T = 50 287 6.

Then. significant oscillations in the electromagnetic torque and speed on motor are observed. the motor speed can not reach to the reference speed). 17. it is observed that the real speed can not follow the reference speed after applying load torque at t = 3 s). a load torque equal to 0. 15 (c)).8 rpm at rotor speed of 500 rpm).(Fig. 315 316 317 318 319 320 321 322 323 Figs. the speed oscillation after applying load torque and at steady state is ~ 8 rpm at rotor speed of 500 rpm but by using proposed modified controller the speed oscillation reduced notably ~ 0.5 rpm at rotor speed of 500 rpm. In Fig. 18 and 19 that the dynamic performance of the proposed system for vector control of faulty three-phase IM is extremely acceptable. Fig. It can be seen from Figs. from Fig. At the same time. when the speed reference changes from 100 rpm to 700 rpm. It is evident from Fig. phase “cs” of stator is opened at t = 1 s and the IM becomes unbalanced. 19. 20 displays simulation results of the proposed modified vector control for faulty IM in the difference values of reference speed.m is applied to the machine. 20 (b) shows the speed error between reference speed and actual speed. 18. the motor is starting in balanced condition.Conclusion In this paper. In the presented technique.m is applied at t = 0. as can be seen in Fig. the modified method (Fig. 17. It can be shown that under fault condition. 19 (c-left). shows that proposed modified controller decreases the torque ripples noticeably. 18 and 19 show the comparison between conventional and modified controller (left: conventional. Moreover. In Fig. the speed oscillation reduced notably ~ 0. In starting. two load torque equal to 0. Simulation illustrates that the conventional RFOC unable to control the faulty IM correctly (In this condition.5 N. the motor is starting in balanced condition. In Fig. 16 demonstrates the conventional RFOC simulated results. right: modified). two methods for modeling (d-q and WFM) and a novel approach for RFOC of a faulty three-phase IM fed by PWM. 339 340 341 342 343 344 345 346 347 7. Simulation results of conventional RFO controller show that the conventional controller is unable to control the IM under open-phase fault properly (e.5 N. a GSA is implanted for tuning PI controllers. by using conventional controller. the similar process is done but this time after open-phase fault. Fig. In Fig.4 s. 20 (a) and (b) that the faulty IM can follow the reference speed without any overshoot and steady-state error.4 s and t = 3 s respectively. by using some changes in the block diagram of conventional 348 349 350 351 352 . 313 314 Fig. Then a phase cut-off is introduced at t = 1 s. It can be seen from the presented results that the dynamic performance of the proposed drive system is highly satisfactory. Moreover.05 s. In this figure a fault cut-off is occurred at the starting. a load torque equal to 0. Fig 20 (c) shows the electromagnetic torque of the motor in the faulty condition. 18. IM is healthy.5 Nm and 1. 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 Fig. 10) is employed. Then a phase cut-off is introduced at t = 0.. 20 (a) shows the reference and actual rotor speed. Simulation results in Fig. three-leg VSI is recommended.g.5 Nm is applied at t = 0. 19. Especially.

625–632. Arellano-Padilla J.6:679–684. It is envisaged that the presented technique for faulty IM control can be used for single-phase IM as well. 1996. Idris NRN. 377 378 379 [9]. IEEE Transaction on Energy Conversion 1995. 1–6. IEEE Energy Conversion Congress and Exposition. 2010. Sumner M. p. CA. 1–8. Portugal. 362 363 364 [4] Jannati M. Evaluation of a vector controlled induction motor drive using the dynamic magnetic circuit model. It can be noted that the faulty IM structure is like single-phase IM structure.38:101–109. Toliyat HA. IEEE International Conference on Industry Applications. rotor bar and end ring faults. White JC. Lipo TA. Tami A. Tehran. Modeling of induction machines for electric drives. we can control the unbalanced or faulty IM. 380 381 [10] Toliyat HA. 374 375 376 [8] Toliyat HA. 365 366 367 [5] Jasim O. p. San Diego. IEEE Transaction on Industrial Applications 2002. Sumner M. IEEE Transaction on Energy Conversion 1991. Bradley K. IEEE Transaction on Energy Conversion 1991. 353 354 355 References 356 [1] Bradley KJ. Modeling and Vector Control of Unbalanced induction motors (faulty three phase or single phase induction motors). Industry Applications Conference.5:618–625. White JC. 368 369 370 [6] Jasim O. Reluctance mesh modelling of induction motors with healthy and faulty rotors. Gerada C. 2008. 208–211. 2003.10:241– 247. Analysis of a concentrated winding induction machine for Adjustable Speed Drive Applications-part I (Motor Analysis). Sumner M. USA. Salam Z. Investigation of induction machine phase open circuit faults using a simpliﬁed equivalent circuit model. Sewell P. UT. Slemon GR. Analysis of a concentrated winding induction machine for adjustable speed drive applications-part II (Motor design and performance analysis). 2012. Vilamoura. 382 383 384 [11] Nandi S. 359 360 361 [3] Jannati M. 385 386 387 . p.RFOC. Salt Lake City. Raleigh. A New Method for Modeling and Vector Control of Unbalanced Induction Motors. In: Proc International Conference on Electrical Machines. Development of a new faulttolerant induction motor control strategy using an enhanced equivalent circuit model. Lipo TA. The 1st IEEE Conference on Power Electronic & Drive Systems & Technologies. p. 3625–3632. Arellano-Padilla J. Lipo TA. Gerada C. Novel frequency-domain-based technique to detect stator interturn faults in induction machines using stator-induced voltages after switch-off.6:685–692. Fallah E.38:101–109. Transient analysis of cage induction machines under stator. 357 358 [2] Gerada C. 371 372 373 [7] Toliyat HA. p. IET Electric Power Applications 2011. NC. IEEE Transaction on Industrial Applications 2002.

Charara A. 423 . Scelba G. Habetler TG.35:1000–1006. On-line stator winding fault detection in inverter fed induction motors by stator current reconstruction. p. IEEE Transaction on Industrial Applications 2013.38:632–637. p. Burton BH. El-Antably A. Zein I. IEEE Transaction on Industrial Electronics 2000. 400 401 402 [17] Eltabach M. IEEE Transaction on Industrial Applications 2002.47:1078–1084. IEEE Transaction on Industrial Applications 1999. p. Oxford science publication 1990. Akin E. Drive Systems and Technologies. 397 398 399 [16] Ghazal M. The 2nd IEEE Conference on Power Electronics. The detection of inter-turn short circuits in the stator windings of operating motors. Tehran. 2011. 409 410 411 [20] Tallam RM. Der Minassians A. Nandi S. 375–380. Consoli A. 2000. The 9th IEEE conference on Electrical machines and drive. p.5:107–121. IEEE Transaction on Industrial Applications 1995. Transient model for induction machines with stator winding turn faults. IEEE Conference in Industry Applications. Neural network based on-line stator winding turn fault detection for induction motors. 163–168. 412 413 414 [21] Tallam RM. 394 395 396 [15] Nandi S. Gritter DJ. 304–309. IEEE Conference in Industry Applications. 391 392 393 [14] Luo X. Transient model for induction machines with stator winding turn faults. 1999. Lipo TA. 421 422 [24] Vas p. 403 404 405 [18] Aydin I. Robust stator winding fault detection in induction motors.42:1201–1208. Liao Y. Kelk HM.49:74–83. 418 419 420 [23] Gaeta A. A Comparison of External and Internal Methods of Signal Spectral Analysis for Broken Rotor Bars Detection in Induction Motors. IEEE Transaction on Industrial Applications 2006. Karakose M. Harley RG. Habetler TG. Habetler TG. 388 389 390 [13] Joksimovic MG. IEEE Transaction On Industrial Electronics 2004. Rome. Poshtan J. Detection of Stator Faults in Induction Machines Using Residual Saturation Harmonics Transient model for induction machines with stator winding turn faults. 415 416 417 [22] Tallam RM. 253–257. Toliyat H. A novel approach for broken rotor bar detection in cage induction motors. Penman J. Vector Control of AC Machines.52:1790– 1799. 2000. Wohrnschimmel R.[12] Milimonfared J. Modelling and Control of Three-Phase PMSMs under Open-Phase Fault. Multiple coupled circuit modeling of induction machines. Canterbury.31:311–318. Rome. Harley RG. A new method for early fault detection and diagnosis of broken rotor bars. 406 407 408 [19] Wolbank TM. Energy Conversion and Management 2011. Toliyat HA. Harley RG.

The ant system: optimization by a colony of cooperating agents. Science 1983. 427 428 429 [27] Abbaszadeh K. Cho JH. Scientia Iranica 2002. Mili Monfared J. Modeling and Control of a Multiphase Induction Machine with Structural Unbalance. 455 456 [39] Ellabib I. Information Sciences 2009. Calamai P. Part I-Machine Modeling and Multi-Dimensional Current Regulation. Packard NH. 433 434 435 [29] Tang KS. Man. Physica D 2 1986. Perelson AS.34:539–557. Nezamabadi-pour H. Modeling and Control of a Multiphase Induction Machine with Structural Unbalance. and Cybernetics–Part B 1996. Saryazdi S. 451 452 [37] Rashedi E. Fallah E. Information Sciences 2006. 444 445 [34] Dorigo M. 424 425 426 [26] Zhao Y. 436 437 [30] Farmer JD.2:187–204. Lipo TA.9:321–327. IEEE Transaction on Energy Conversion 1996. Man. Abraham A. GSA: A Gravitational Search Algorithm. Kwong S. A study of particle swarm optimization particle trajectories. 177:3918–3937. Swarm intelligence 2007. Man KF. Modeling and Simulation of Dual Three-Phase Induction Machine Under Fault Conditions and Proposal of a New Vector Control Approach for Torque Oscillation Reduction. Central force optimization: a new nature inspired computational framework for multidimensional search and optimization. Exchange strategies for multiple ant colony system.13:22–37. A proof of convergence for ant algorithms. IEEE Transaction on Energy Conversion 1996. IEEE Signal Processing Magazine 1996.220:671–680. Maniezzo V. Vecchi MP. Passino KM.1:33–57. IEEE Transaction on Systems. A hybrid genetic algorithm and bacterial foraging approach for global optimization.160:267–279. 442 443 [33] Kirkpatrick S. Kennedy J.179:2232–2248. 457 458 . Lipo TA. Colorni A. 430 431 432 [28] Formato RA. 449 450 [36] Kim DH. Basir O.3:570–577.26:29–41. adaptation and machine learning. 446 447 448 [35] Gazi V.176:937–971. Blackwell T. Information Sciences 2004. Studies in Computational Intelligence 2008.177:1248–1264. 440 441 [32] Poli R. 438 439 [31] Bergh FVD. Optimization by simulated annealing. Stability analysis of social foraging swarms. He Q. Engelbrecht AP. Gelatto CD. Information Sciences 2007.3:578–584. 453 454 [38] Badr A. IEEE Transaction on Systems.129:221–238. Part II-Field-Oriented Control and Experimental Verification. Particle swarm optimization. Genetic algorithms and their applications. Fahmy A. and Cybernetics–Part B 2004.[25] Zhao Y. The immune system. Information Sciences 2007.

Fault Tolerant Field Oriented Control of the Induction Motor for Loss of One Inverter Phase. Bulkan S. A synergy of differential evolution and bacterial foraging algorithm for global optimization. The 32nd IEEE Conference on Industrial Electronics. 472 473 474 475 [46] Liu H. Kingsley C. Advanced Engineering Informatics 2005. A comparative study of differential evolution. Information Sciences 2008.178:4421–4433. 467 468 [44] Uysall O.4:159–175. Cano JR. Thomsen R. 1980–1987. Bandyopadhyay S. Information Sciences 2007. Clerc M. and evolutionary algorithms on numerical benchmark problems. McGraw-Hill 2003. 4: Turn function of the first rotor bar for simulated motor 489 Fig. 478 479 480 [48] Fitzgerald AE. Herrera F. Multi-objective particle swarm optimization with time variant inertia and acceleration coefficients.17:607–626. 2004. 459 460 461 [41] Lozano M. 482 483 484 Figure Caption: 485 Fig. Umans SD. Neural Network World 2007. Abraham A. Dasgupta S. Applied Mathematics and Computation 2008. p. Paris.177:5033–5049. Improving genetic algorithms’ performance by local search for continuous function optimization. Das S. Replacement strategies to preserve useful diversity in steady-state genetic algorithms.7:1019–1026. 469 470 471 [45] Vesterstrom J. Oregon (EEUU). 1: Faulty IM drive 486 Fig. 464 465 466 [43] Elbeltagi E. 462 463 [42] Tripathi PK. 2: Rotor and stator winding's flux axes under open-phase fault 487 Fig. Electric Machinery.196:309–317. 3: Turn function of stator phases for simulated motor 488 Fig. Applied Soft Computing Journal 2007. Portland. Abraham A. IEEE Congress on Evolutionary Computation. 6: Winding function of the first rotor bar for simulated motor 491 . particle swarm optimization.19:43–53.[40] Hamzacebi C. 481 [49] Saleh A. Pacas M. 5: Winding function of stator phases for simulated motor 490 Fig. Comparison of genetic algorithm and particle swarm optimization for bicriteria permutation flowshop scheduling problem. Chaotic dynamic characteristics in swarm intelligence. 476 477 [47] Biswas A. International Journal of Computational Intelligence Research 2008. Shaltout A. Hegazy T. 817–822. p. 2006. Comparison among five evolutionary-based optimization algorithms. Grierson D. Pal SK.

14: Simulation results of the conventional vector controller. 13: Simulation results of the modified vector controller. (b) Stator a-axis current. (a) Rotor a-axis current.Fig. (d) Electromagnetic torque 511 512 Fig. 9: Simplified single-phase IM equivalent circuit 494 Fig. (b) Speed error. 11: Principle of GSA [37] 496 Fig. (a) Rotor a-axis current. (b) Stator a-axis current. (a) Rotor a-axis current. (b) Stator a-axis current. backward and stationary reference frame 492 Fig. (b) Stator a-axis current. (b) Stator a-axis current. 19: Simulation results of the conventional (left) and modified (right) controller. 16: Simulation results of the conventional vector controller. (d) Electromagnetic torque 497 498 Fig. (a) Rotor a-axis current. (b) Stator a-axis current. 10: Proposed block diagram for faulty IM 495 Fig. (b) Stator a-axis current. (c) Speed. (d) Electromagnetic torque 505 506 Fig. 7: Forward. (c) Speed. (d) Electromagnetic torque 509 510 Fig. 20: Simulation results of the modified controller in the difference values of reference speed. (c) Speed. (d) Electromagnetic torque 503 504 Fig. (c) Electromagnetic torque 513 514 Table Caption 515 Table (1): Comparison between d-q model of the balanced and faulty three-phase IM equations Table (2): Comparison between stator voltage equations of conventional RFOC and modified RFOC Table (3): Comparison between two vector control methods 516 517 518 519 520 521 522 523 524 525 . 17: Simulation results of the modified vector controller. (d) Electromagnetic torque 507 508 Fig. (c) Speed. (b) Stator a-axis current. (a) Rotor a-axis current. (c) Speed. (a) Rotor a-axis current. (a) Rotor a-axis current. (d) Electromagnetic torque 501 502 Fig. (c) Speed. 15: Simulation results of the modified vector controller. (c) Speed. 18: Simulation results of the conventional (left) and modified (right) controller. (a) Speed. (d) Electromagnetic torque 499 500 Fig. 8: Single-phase IM equivalent circuit 493 Fig. (a) Rotor a-axis current. (c) Speed. 12: Simulation results of the conventional vector controller.

[RR] and [LRR]. the equations of faulty motor with different stator windings in a stationary reference frame (superscript “s”) can be expressed as equations (4-9): The matrixes Rs. Lss. [iabS]. Obtaining equation of faulty motor in the d-q frame Applying the transformation vectors ([TS] and [Tr]). stator and rotor voltages and fluxes can be written as following equations: By simplifying above equation. Rr. [LSR] and [LRS] have the similar formation to the balanced IM. [RS]. are identical to the balanced three-phase IM. It is because the structure of rotor has not changed. [abS]. [LSS]. Lsr and Lrr in the WFM are defined as follows: Rb 0 2Rb Re R 2Rb Re Rb b Rr 0 0 0 Rb 0 0 0 0 2Rb Re Rb 536 537 538 539 540 541 542 543 544 545 546 547 548 549 Rb 0 Rb 2Rb Re 550 . [vabcR]. flux and torque equations for faulty three-phase IM in the “abc” frame can be written as following equations: Stator and rotor voltage equations: 527 528 529 530 21 vabS RS22 31 32 vabcR 0 21 1 d 2abS 0 23 iabS 3 1 3 1 33 RR iabcR dt abcR 531 532 Stator and rotor flux equations: 21 abS 31 abcR L L 21 SS 23T SR L i L i 23 SR 33 RR 21 abS 31 abcR 533 534 Torque equation: e Pole L [iabS ]T [ SR ][iabcR ] 2 r 535 where. The [vabS]. The only different in these matrixes are the row and column for the phase “cs” is removed.Appendix A 526 Obtaining equation of faulty motor in the “abc” frame: Voltage. [iabcR]. [abcR].

Lmr is the magnetizing inductance of each rotor loop. [Rs] is a diagonal m by m consists of resistances of each coil. I 33 rs 0 0 0 rs 0 0 0 rs s s L11 L12 Lss Ls21 Ls22 Ls31 Ls32 s L13 s L23 Ls33 sr sr L11 L12 L1srm Lsr Lsr21 Lsr22 Lsr2m Lsr31 Lsr32 Lsr3m Lr1r2 Lb Lr1r3 Lr1rm Lb Lmr 2Lb Le L L Lmr 2Lb Le Lr2r3 Lb Lr2rm r2r1 b Lr3r1 Lr3r2 Lb Lmr 2Lb Le Lr3rm Lrr Lr r Lb L L L 2 L L rmr2 rm r3 mr b e m1 In these matrixes. The matrix [Lss] is a symmetric m by m matrix. [Rr] is n+1 by n+1 symmetric where. Lb is the rotor bar leakage inductance.Rs rs . Rb is the rotor bar resistance and Re is the end ring segment resistance. Appendix B 551 552 553 554 555 556 557 558 559 560 561 562 Equations of faulty IM by using proposed rotational transformations: Torque equation: Rotor voltage equations: 563 Stator voltage equations: 564 . The mutual inductance matrix [Lsr] is an m by n matrix comprised of the mutual inductances between the stator coils and the rotor loops. Le is the rotor end ring leakage inductance and Lri rk is the mutual inductance between two rotor loops.

0038 kg.m2. f = 50 Hz.15 .851 H 571 572 573 . Lms = 0. rs = 20. Power = 475 W. rr = 19.6 . J = 0.565 where: 566 567 Appendix C 568 569 570 The ratings and parameters of three-phase IM are: Voltage: 125 V.0814 H. Llr = Lls = 0. of poles = 4. No.

- Design of Hybrid Controller for Direct Torque Control of Induction Motor DriveUploaded byIRJET Journal
- ijee98Uploaded byMohamed Allaban
- A Microcomputer-based Induction Motor Drive_1999Uploaded byThyago Vasconcelos
- 01157850.pdfUploaded byJorge Guerrero
- Bai 2013Uploaded byBobby Rinaldi
- Rotor Ident 11Uploaded bymakroum
- Final 1Uploaded byChidanand Hebbale
- Inverse FuzzyUploaded byDaniel Gutierrez
- PaperUploaded bymeghraj01
- A Robust RF-MRAS Based Speed Estimator Using Neural Network as a Reference Model for Sensor-Less Vector Controlled IM DrivesUploaded byAlexander Decker
- DocumentUploaded byA Rana
- Editing-Sensorless-Flux-Vector-Control-of-Induction-Motor.docxUploaded byHanny Prastya Hariyadi
- DYNAMIC MODEL OF INDUCTION MOTORS FOR VECTOR CONTROLUploaded byDiptarka Moitra
- 16_msivasathyanarayana_finalpaper--IISTE research paperUploaded byiiste
- 01375643Uploaded bySushant Kumar
- 1639014Uploaded byRoshindas Pudhiya Purayil
- fulltext_140Uploaded byinfodotz
- Motor and Starter TrainingUploaded byGavrila George
- WEG INVERTER MOTOR _ANTECHUploaded byOG
- Yaskawa A1000 Catalog.pdfUploaded byLê Phụng
- Motor ControlUploaded bycorin_o
- New Microsoft Office Word Document (2)Uploaded bySuryaprasad Samsani
- Main PaperUploaded bysreeragks1989
- LCR meterUploaded byhungphuc2010
- Unit 3.pdfUploaded bymanoj kumar
- Induction MotorUploaded byshijub_001
- SE Electrical SyllabusUploaded byyogesh_b_k
- Servo and Robot DrivesUploaded byDaniloT.HungaroMicheletti
- IdentificationUploaded byramiyamin
- SrikantaMurthy Umn 0130M 17368Uploaded byHồ Minh Hới

- 4_types of dc chopper.pdfUploaded byPrem Kumar
- MIT9_00SCF11_text.pdfUploaded byPrem Kumar
- 1_Architecture of 8085.pdfUploaded byPrem Kumar
- co design.pptUploaded byPrem Kumar
- CHAPTER 5.pptxUploaded byPrem Kumar
- PSY_chapter__1.pdfUploaded byPrem Kumar
- 77442_92066v00_motor-control-with-arduino-a-case-study-in-design.pdfUploaded byPrem Kumar
- Chapter 4 bps.pptxUploaded byPrem Kumar
- Dream-Psychology-Sigmund-FreudPsychoanalysis-for-Beginners.pdfUploaded byPrem Kumar
- 3_three phase converter.pdfUploaded byPrem Kumar
- Chapter3PPT.pptxUploaded byPrem Kumar
- chap4.pptxUploaded byPrem Kumar
- using-arduino-with-matlab-and-simulink.pdfUploaded byPrem Kumar
- 5_AC motor starter.pdfUploaded byPrem Kumar
- 2_AC-DC Converter.pdfUploaded byPrem Kumar
- Grid.pptUploaded byPrem Kumar
- anfisUploaded byPrem Kumar
- ELK-1303-143_manuscript_1.pdfUploaded byPrem Kumar
- MECH II TO VIIIUploaded byPrem Kumar
- ELK-1212-31_manuscript_1.pdfUploaded byPrem Kumar
- ELK-1303-143_manuscript_1.pdfUploaded byPrem Kumar
- ELK-1305-114_manuscript_2.pdfUploaded byPrem Kumar
- ItaUploaded byPrem Kumar
- ELK-1212-31_manuscript_1.pdfUploaded byPrem Kumar
- Cognitive Psychology.pdfUploaded byPrem Kumar
- ELK-1212-31_manuscript_1.pdfUploaded byPrem Kumar
- pwmUploaded byPrem Kumar
- ELK-1305-114_manuscript_2.pdfUploaded byPrem Kumar
- ELK-1108-38_manuscript_2.pdfUploaded byPrem Kumar