Closed-Loop Bandwidth Impact on MVSA for Rotor Broken Bar Diagnosis in IRFOC Double Squirrel Cage Induction Motor Drives

Transcription

1 Cloed-Loop Bandwdth Impact on MVSA for Rotor Broken Bar Dagno n IRFOC Double Squrrel Cage Inducton Motor Drve Y. Grtl *, ***, A. O. D Tommao **, R. Mcel **, F. Flppett *, and C. Ro * * Unverty of Bologna, (Italy) Dpartmento d Ingegnera dell Energa Elettrca e dell Informazone «Guglelmo Marcon». E-mal: yaer.grtl@unbo.t, forenzo.flppett@unbo.t, claudo.ro@unbo.t. ** Unverty of Palermo, (Italy) Department of Electrcal, Electronc and Telecommuncaton Engneerng, Chemcal Technologe, Automatc and Mathematcal Model. E-mal: dtommao@deet.unpa.t, roaro.mcel@unpa.t. *** Unverty of Tun El Manar, (Tuna) LARA-Department of Electrcal Engneerng, Natonal Engneerng School of Tun. E-mal: yaer.grtl@et.rnu.tn. Abtract Th paper nvetgate the detectablty of rotor broken bar n ndrect rotor flux orented control (IRFOC) for varable peed double cage nducton motor drve, ung vbraton gnature analy technque. The Impact of the cloed loop control ytem cannot be neglected when the detecton of rotor aymmetre n the machne baed on the gnature analy of electrcal or mechancal varable. Therefore, the nvetgaton of rotor fault component for dfferent bandwdth of cloed-loop regulator neceary to evaluate t relevance n the above lted varable. Th paper nvetgate the mpact of the control ytem on relevance of the fault component computed from axal and radal vbraton gnal. Expermental reult how the valdty of th mpact, and more pecfcally, the bandwdth PI regulator. Due to t relevance, axal vbraton analy how a more robut fault gnature, under the control mpact, n eparatng healthy from rotor bar breakage n double qurrel cage nducton motor. Index Term-- AC Machne; Condton montorng; double cage rotor; rotor broken bar; nducton motor drve. I. INTRODUCTION In the lat few year, new dagnotc procedure have been nvetgated manly focung on renewable energy ource and power converon ue. Improvng the power converon effcency ha quckly become the key target even f contratng wth ytem performance. New challenge are thu preng both academc and ndutral reearch toward nnovatve oluton [1-7]. Effcent energy conumpton a key factor to Europe ambtou goal for utanable development and actvte related to ar polluton and clmate hftng. In order to tackle the ncreang electrcty demand, a number of oluton for effcent energy conumpton, generaton of energy from renewable ource, and new power dtrbuton bune model for actve energy control have been condered. Fuel cell baed hybrd power upply ytem for houehold applance are actually wdely nvetgated. Sophtcated control algorthm are uually mplemented to meet hydrogen avng and effcency requrement. Great advantage on ytem fault detecton, preventon and recovery are brought by accurate ytem modellng, ncludng the energy ource telf [8]. ey beneft are brought by accurate meaurement et up and method [11] Snce photovoltac (PV) ntalled power ncreae all over the world, olar technology now trongly preent n the electrcty market, and t cannot be een only a a von of the future. PV generaton ytem are actually avalable n dfferent power ze coverng the range from dometc applcaton to large power generaton plant [12-14]. Many dfferent mprovement have been propoed and expermented to enhance effcency not only n energy converon (PV feld) but alo n energy management (power converter). In recent tme, mult-level nverter uch a H brdge mult-cell cacaded nverter, Neutral Pont Clamped (NPC) and Actve Pont Clamped (APC) have alo been ued for a more effcent and effectve PV power converon [15-16]. Many nverter topologe have been patented, wthn the market development, to olve ome common problem of the PV plant uch a leakage current due to paratc capactance of the panel and effcency loe due to the tranfer of reactve energy between the grd flter and the DC bu capactor. Moreover, ome partcular topologe have been conceved of for the fault tolerant operaton of PV plant. In renewable energy ource tate of the art an mportant role now played from ea electrc energe generator, between one of the mot nteretng the drect generaton wth a tubular lnear generator. In the lat fve year, the number of operatng machne etmated to 16.1 bllon n 2011, wth growth of about 50% [17]. Nowaday, the number of electrcal machne ntegrated n cloed-loop drve ha ganed ground wthn /13/$ IEEE 529

2 mot ndutral area. Invetgaton on dfferent falure mode n nducton motor have revealed that 19% of the overall motor fault are related to the rotor part [18,19]. A detaled analy of th category of fault can be found n [20]. Double cage rotor are characterzed by t large outer (tartng) bar retance and large nner (runnng) bar leakage nductance. Snce double cage motor are ued for loaded tartup, the outer cage mut handle the large tartng current for the long acceleraton tme, wth lmted path for heat dpaton [21]. Cyclc electromagnetc force, thermal tre, envronmental tre, and mechancal tre aocated wth the large tartng current, make the outer cage more vulnerable to fatgue falure. Dfferent technque are ued to provde accurate non-nvave on-lne condton montorng of nducton motor. Intead of tator current, mechancal or electrcal quantte were ued a reference gnal to detect the preence of rotor fault. A new technque preented n [22], where the electromagnetc torque magntude computed n dfferent way a a functon of rotor poton, to detect the preence of rotor fault. Reference [23] and [24] are devoted to detecton of broken bar by external magnetc feld analy. Motor vbraton gnature analy (MVSA) wa nvetgated n [25], howng the effectvene of th technque n dagnong rotor broken bar n lne-connected motor. In order to dcern cae n whch the preence of nter bar current decreae the entvty of the motor current gnature analy (MCSA), motor axal vbraton gnature analy (MAVSA) and motor radal vbraton gnature analy (MRVSA) were nvetgated n [25,26]. Recently, th ue wa nvetgated for lne-fed double cage motor, operatng under teady-tate condton n [26] and more recently n [27,28]. On the other hand, the rotor current manly flow n the ymmetrc nner cage under teady tate operatng condton, and fault gnature nentve to outer cage damage. A combned ue of current and vbraton analy wa developed, by correlatng the gnal pectra to enhance broken bar detecton ablty under loaded and unloaded operatng condton of the motor n [29]. It hould be noted that for the above lted contrbuton, referred to the detecton of broken bar baed on vbraton gnature analy, the applcaton are condered for lne-connected motor. On the other hand, t wa hown that n cloed loop drve the control telf affect the behavor of electrcal varable and o new dagnotc procedure are neceary for machne montorng. More pecfcally, the tunng of the peed and current controller ued n cloed-loop drve ha a conderable mpact, whch can compenate the ocllaton nduced on the dfferent varable n preence of rotor fault condton [28-36]. Modulated gnal were nvetgated n teady-tate condton and then under-tme varyng condton repectvely n [30] and [31] to overcome the compenatng effect of the control ytem on current harmonc to dagnoe rotor fault. The meaured drect and quadrature current component were propoed n [32] and [33] repectvely for the detecton of rotor fault. In [34], an exhautve nvetgaton of the mpact of the control trategy and even the tunng of the peed regulator wa condered a potental crtera for the electon of the approprate motor fault dagno technque. Vrtual magnetzng current propoed n [36] to detect and quantfy rotor broken bar n nducton motor. In general, when fault occur n the machne, the typcal current fault-harmonc mght become le vble due to the compenatng acton of the control ytem. Therefore, the relevance of the typcal fault component ued from MVSA could be affected by the bandwdth of current regulator. For th reaon, the entvty of the MVSA nvetgated n repect to the bandwdth varaton for double cage nducton motor broken bar. Expermental reult how clearly that th mpact not neglgble, and the dagnotc procedure baed on MVSA n cloed-loop operatng condton mut be condered n order to perform an effectve rotor broken bar detecton. II. ROTOR FAULT FREQUENCIES PROPAGATION Lke any other rotatng electrcal machne, the double cage nducton motor ubjected to electromagnetc and mechancal force ymmetrcally reparttoned. In healthy condton only the fundamental frequency f ext n tator current (f : upply frequency). If the rotor part damaged, the rotor ymmetry of the machne lot producng a revere rotatng magnetc feld related to an nvere equence component at frequency f. Th nvere equence reflected on the tator de, producng the frequency (12)f. Thee frequency component generate electromagnetc and mechancal nteracton between tator and rotor part. Conequently torque and peed rpple effect are generated at frequency 2f, whch modulate the rotatng magnetc flux [36]. Frequency Propagaton Rotor freq. ±3f ±f Stator freq. (1+2)f (1-2)f f Mechancal peed freq. 2f 1/f ±1/f* Tme-Frequence Propagaton 1/(1-2)f 1/2f 1/(1+2)f Tme Propagaton ±1/3f* Rotor perod Stator perod Mechancal peed perod Fg. 1. Block dagram of the control cheme for a typcal ndrect feldorented control nducton motor drve. Th modulaton produce two current component,.e., an addtonal left-de component at (12)f and a rght de component at (1+2)f. Followng th nteracton proce, the frequency content of the tator current how ere of fault component at the followng frequence chan: fkr 1 2 k f k 1,2,3,... (1) 530

3 More pecfcally, for large or double qurrel-cage motor, a long a the contact mpedance between the rotor bar and ron core mall or the copper bar are drectly nerted nto the lamnated ron lot, the broken bar no longer a phycal condton enurng an open crcut and nter bar or cro-path current can flow. A a conequence, thee nter bar tranvere current nteract wth the radal tator flux denty, generatng axal force. Thee fact lead manly to the preence of two chan of fault component: fmec 2kf k1,2,3, kf k 1,2,3,... vb_1 f kr (2) vb_ 2 f kr (3) both n radal and axal vbraton drecton (f mec denote the mechancal peed of the motor) [30]. In next ecton, the focu wll be excluvely on trackng the detectablty of the mot relevant fault component (f mec -2f) of the harmonc chan (f mec ±2kf) k=1,2,3, n frequency doman, n preence of the control dampng effect. III. IMPLEMENTED CONTROL STRATEGY: EXPERIMENTAL SETUP Indrect rotor feld orented control one of the mot wdely ued control trategy for nducton motor drve n ndutral applcaton. In the followng, a decrpton of the nvetgated control trategy, and t expermental mplementaton upplyng a double cage nducton motor under healthy and rotor broken bar. A. Control Strategy Decrpton A llutrated by the block cheme n Fg. 2, the double cage nducton motor control ytem baed on two cacaded loop. The outer loop dedcated to the peed control. Wherea the nner one related to the control of the drect and quadrature tator current component. The block contanng D and D-1 repreent repectvely the Clarke and the nvere Clarke tranformaton from a three phae to a two phae d-q ytem. A clacal PI regulator ued to generate the reference value of the torque, whch produce the current component q *. The two other PI regulator are dedcated to the control of the d-q component of the tator current. Snce the control ymmetry of the d-q current component, a llutrated by Fg. 3, the PI regulator have the ame parameter for the d-ax and the q-ax loop. Conderng a perfect decouplng of the two ax, the drect and quadrature v d * and v q * voltage equaton are: v v * d * q R d R q L L d dt d dt where R the tator retance, L the tator nductance, and the leakage coeffcent. Ung a zero pole cancellaton and conderng for the PI regulator the followng tranfer functon: PI ( ) p (6) d q (4) (5) The parameter of the regulator can be derved a follow: p 2Bp L (7) 2 B R (8) c * e q* Speed PI reference q Speed regulator Reference generaton rd q* d* Flux reference generaton d/dt q d 1/ p L p L p d* d (a) p p PI PI vq* vd* D -1 D R R r P W M 1 L 1 L a b c E dc I.M. vq* q d vd* (b) Fg. 2. Block dagram of a typcal ndrect feld-orented control nducton motor drve; (a) Implemented IRFOC drve; (b) current control loop for PI controller degn. Beng B p the bandwdth of tator current cloed-loop. In the next ecton, the mpact of the above bandwdth on rotor bar breakage detectablty baed on axal and radal vbraton analy wll be evdenced. B. Expermental Setup Decrpton A complete et of expermental tet have been carred out. Fg. 3 how detal of the expermental tetbed. For th purpoe, two double cage nducton motor are avalable; one healthy, and the econd wth a drlled broken bar (the bar wa completely dconnected from the common end-rng). Photo of the ued healthy and faulty double cage rotor are hown n Fg. 3. The charactertc of the double cage nducton motor ued for experment are preented n Table I. A PWM VSI converter ued to control the motor. A dspace ytem baed on the DS1103 control board ha been ued to mplement the ndrect rotor feld orented control hown n Fg. 2. Two pezoelectrc accelerometer Brüel & jær model 4507 B 005, were mounted for meaurng axal and radal vbraton of the core motor. A NEXUS 2693 model a gnal condtoner ued 531

4 for vbraton gnal. The control board alo ued for data acquton and data proceng. The expermental tet have been performed by operatng the machne under full load operatng condton. (a) (b) (a) (b) (c) Fg. 3. Expermental Tet-bench detal ;(b) accelerometer poton, and the current probe; (c) healthy and drlled rotor bar. TABLE I INDUCTION MOTOR PARAMETERS Data Value Rated Power kw 5.5 Rated tator voltage V 400 Rated current A 13 Rated frequency Hz 50 Rated peed rpm 2870 Rotor dameter mm 110 Axal length of the rotor mm 90 Ar gap length mm 0.5 IV. ROTOR BAR BREAAGE DETECTION: CONTROL IMPACT ON MVSA In th ecton, the detectablty of rotor bar breakage ung MVSA, preented and dcued. Fg. 4. how the MAVSA pectra ued from expermental reult, under healthy and rotor bar breakage. The magntude are normalzed to the maxmum ampltude of the harmonc component at frequency 6f. The pectrum under healthy machne condered a reference n comparon to the faulty cae. Axal vbraton gnal pectra cloe to the mechancal frequency f mec are depcted n Fg. 4 for full load opearatng condton; under healthy (Red) and rotor bar breakage (Blue). A t can be een, under faulty condton the left de component (f mec -2f) ncreae n a gnfcant way n comparon to the healthy cae. More pecfcally, the left de component (f mec -2f) how a gnfcant ncreae from -41,25 db under healthy condton, to -32,36dB for the bar breakage motor. Let u now focu on the contrbuton of the fault component (f mec -2f) ued from radal vbraton analy. MRVSA. Radal vbraton gnal pectra cloe to the mechancal frequency f mec are depcted n Fg. 5 for the ame above full load opearatng condton; under healthy (Red) and rotor bar breakage (Blue). A t can be een, under faulty condton the left de component (f mec -2f) ncreae n a gnfcant way n comparon to the healthy cae. More pecfcally, the left de component (f mec -2f) how an ncreae from -42,03dB under healthy condton, to -33,20dB for the bar breakage motor. By comparng the contrbuton of the fault component (f mec -2f) ued from MAVSA and MRVSA, under healthy and faulty condton, t obvou that the relevance of th component more mportant from axal than from radal vbraton gnal. Th reult expected, accordng to the prevou nvetgaton [27,28], where the contrbuton of the fault component (f mec -2f) from MAVSA and MRVSA, wa nvetgated n lne-fed condton, under healthy and bar breakage condton. Once agan, the axal vbraton analy a relable technque for the detecton of rotor bar breakage n double qurrel cage nducton motor ntegrated n cloed-loop ytem. But t worth notng that the varaton n magntude conderably reduced under cloed-loop operatng condton when compared to the lne-fed operatng condton. Ampltude (db) -70 Fg. 4. Axal vbraton pectra at full load operatng condton; healthy (Red), and rotor bar breakage (Blue). Ampltude (db) f mec -2f f mec -2f -70 Fg. 5. Radal vbraton pectra at full load operatng condton; healthy (Red), and rotor bar breakage (Blue). In order to evaluate the mpact of the control cloedloop bandwdth mpact, dfferent expermental tet were performed wth dfferent value for the current PI controller parameter, correpondng to dfferent cloedloop bandwdth tartng from 200Hz to 70Hz. In Fg. 6, the expermental reult, ued from MAVSA, for rotor bar breakage condton, and under dfferent bandwdth controller parameter are preented. It evdent that the contrbuton, of the fault component of nteret, 532

5 damped by the acton of the control ytem. Obervng the zoomed vew of the magntude fault component, t clear that the lower magntude regtered for the larget bandwdth current PI controller (200Hz). On the other hand, mlar analy baed on radal vbraton analy wa performed. In Fg. 7, the expermental reult, ued from MRVSA, for rotor bar breakage condton, and under the ame dfferent bandwdth controller parameter are preented. The contrbuton, of the fault component of nteret, damped by the acton of the control ytem. The ame obervaton notced for the axal vbraton gnature analy are vald alo for MRVSA. Obervng the zoomed vew preented n Fg. 6 and Fg. 7 correpondng to MAVSA and MRVSA repectvely, t can be een that the control mpact, on rotor bar breakage entvty, more mportant on MRVSA than MAVSA. Ampltude (db) B =70Hz -15 p B =90Hz p B p =200Hz Fg. 6. Axal vbraton pectra at full load operatng condton; bandwdth mpact on the (f mec-2f) component under rotor bar breakage. Ampltude (db) B =70Hz -15 p B =90Hz p B p =200Hz Fg. 7. Radal vbraton pectra at full load operatng condton; bandwdth mpact on the (f mec-2f) component under rotor bar breakage. V. CONCLUSION The clacal motor current gnature analy technque may fal due to the preence of nter bar current, whch reduce the ampltude of the tracked fault component n frequency doman. Vbraton analy may not only detect the extence of nter bar current under broken bar falure, but alo enhance the detectablty of th type of fault by conderng clacal gnal proceng technque. For mall nducton motor, the MCSA dagnotc procedure tll effectve. But, wth hgh power or double cage motor, lke the teted one, the combned ue of current and vbraton analy recommended n order to avod erroneou dagno. A proved n prevou contrbuton [27,28], the gnature ued from axal vbraton analy more relevant, n comparon to the current and radal vbraton analy reult n lne-fed nducton motor. In th paper, the robutne of the fault gnature ued from axal vbraton n cloed-loop operaton, n comparon to the radal vbraton one, evdenced. Effectvely, due to the mportant relevance of the fault component ued from axal vbraton, th ue make the detectablty le entve to the control parameter. Trackng thee fault component n Tme-Frequency doman our man challenge n the future to guaranty a relable rotor broken bar fault gnature n a double cage nducton motor n cloed-loop operatng condton. ACNOWLEDGMENT Th work wa realzed wth the contrbuton of SDES Laboratory - UNINETLAB - Unverty of Palermo. REFERENCES [1] V. Bocano, P. Lvrer, F. Marno, and M. Mner, Lnear-non-lnear Dgtal Control for DC-DC Converter wth Fat Tranent Repone, ACTAPRESS, Internatonal Journal of Power and Energy Source, 2009, vol. 29, pp Do: /Journal [2] G.M. D Bla, V. Bocano, P. Lvrer, F. Marno, and M. Mner, A Novel Lnear-Non-Lnear Dgtal Control for DC/DC Converter wth Fat Tranent Repone, IEEE Appled Power Electronc Conference, APEC2006, USA pp DOI: /APEC [3] V. Bocano, M. Gaeta, G. Cappon, F. Marno, "Nonlnear dgtal control mprovng tranent repone: Degn and tet on a multphae VRM," Sympoum on Power Electronc Electrcal Drve Automaton and Moton (SPEEDAM), June, 2010, pp [4] V. Bocano, P. Lvrer, F. Marno, and M. Mner, Current-Senng Technque for Current-Mode Controlled Voltage Regulator Module. Mcroelectronc Journal, Elever, vol.39, no.12, Dec. 2008, pp [5] V. Bocano and G. Cappon, A Hgh-Effcency, Low- Cot Soluton for On-Board Power Converter, Advance n Power Electronc, Hyndaw, vol. 2012, pp do: /2012/259756, Sept [6] A. Danner, R. Rzzo: An overvew of Power Electronc Tranformer: Control tratege and topologe, Sympoum on Power Electronc, Electrcal Drve, Automaton and Moton SPEEDAM 2012, Sorrento, June 20-22, 2012, pp [7] L. Pegar, R. Rzzo, P. Trcol: A Comparon between Lne-Start Synchronou Machne and Inducton Machne n Dtrbuted Generaton, Prezglad Elektrotechnczny (Electrcal Revew), Vol. 88 n. 5b, 2012, pp [8] V. Bocano, G. Cappon, and F. Marno, Expermental tet on a fuel cell-upercapactor hybrd power upply for a dgtal tll camera IEEE Unverte Power Engneerng Conference, UPEC 09, Scotland, 2009, pp.1-5. [9] V. Bocano, G. Cappon, P.Lvrer, and F. Marno, Fuel Cell Modellng for Power Supply Sytem Degn. IEEE Internatonal Workhop on Control and Modelng for Power Electronc, COMPEL 08, Zurch, pp

6 [10] V.Bocano, R. Mcel, G. Cappon,D. Caade Fuel cell Modellng and Tet:expermental valdaton of model accuracy, 4th IEEE Internatonal Conference on Power Engneerng,Energy and Electrcal Drve 2013, POWERENG 2013, May 2013, Turkey. pp [11] C. Spataro: ADC Baed Meaurement: a Common Ba for the Uncertanty Etmaton, Proc. of 17th IMEO - TC4 Sympoum. oce, Slovaka. Sept. 08, [12] R. Candela, V. D Do, E. Rva Saneverno, and P. Romano, Reconfguraton technque of partal haded PV ytem for the maxmzaton of electrcal energy producton, Internatonal Conference on Clean Electrcal Power, ICCEP 07, 2007, pp [13] G. Cpran, V. D Do, D. La Caca, R. Mcel, and R. Rzzo, A novel approach for parameter determnaton n four lumped PV parametrc model wth operatve range evaluaton, Internatonal Revew of Electrcal Engneerng (I.R.E.E.), vol. 8, Apr [14] V. D Do, R. Mcel, C. Rando, and G. Zzzo, Dynamc photovoltac generator: Techncal apect and economcal valuaton, Sympoum on Power Electronc Electrcal Drve Automaton and Moton SPEEDAM, 2010, pp [15] D Tommao A O, Genduo F, Mcel R, Rcco Galluzzo G. Computer Aded Optmzaton Va Smulaton Tool of Energy Generaton Sytem Wth Unveral Small Wnd Turbne. IEEE Internatonal Sympoum on Power Electronc for Dtrbuted Generaton Sytem (PEDG) Aalborg, Denmark, June , p [16] Cecat C., Genduo F., Mcel, R., Rcco Galluzzo G., A utable control technque for fault-tolerant converter n dtrbuted Generaton,IEEE Internatonal Sympoum on Indutral Electronc ISIE 2011, Gdank, 27 June 2011, pp ISBN: [17] H. A. Tolyat, S. Nand, S. Cho, and H. Mehgn-elk, Electrc Machne, Modelng, condton montorng and Fault dagno. CRC Pre, [18] V. D Do, G. Cpran, R. Mcel, R. Rzzo, Degn crtera of tubular lnear nducton motor and generator: a prototype realzaton and t characterzaton, Leonardo Electronc Journal of Practce and Technologe, vol. 12, [19] A.H. Bonnett, C. Yung, "Increaed Effcency Veru Increaed Relablty", IEEE Indutry Applcaton Magazne, Vol. 14, Iue 1, Jan./Feb [20] F. Flppett, A. Belln, G. A. Capolno, " Condton Montorng and Dagno of Rotor Fault n Inducton Machne: State of Art and Future Perpectve", IEEE- Workhop on electrcal machne, degn, control and dagno IEEE-WEMDCD 13, Par, March [21] J. Mroz, "Temperature feld of a double qurrel-cage motor durng tartup," Electrc Power Applcaton, IEE Proceedng -, vol. 152, no. 6, pp , nov [22] R. S. Weer, C. ral, and F. Prker, "The Venna nducton machne montorng method; on the mpact of the feld orented control tructure on real operatonal behavor of a faulty machne," IECON, 1998, vol. 3, pp [23] A. Ceban, R. Puca, and R. Romary, Study of rotor fault n nducton motor ung external magnetc feld analy, IEEE Tranacton on Indutral Electronc, vol. 59, no. 5, pp , may [24] A. Yazd, H. Henao, G. A. Capolno, M. Artol, F. Flppett, and D. Caade, "Flux gnature analy: An alternatve method for the fault dagno of nducton machne," IEEE PowerTech, St. Peterburg, Rua, Jun. 2005, Paper n 424, CD-ROM. [25] C. Concar, G. Francechn, and C. Taon, "Dfferental Dagno Baed on Multvarable Montorng to Ae Inducton Machne Rotor Condton", IEEE Tran. on Indutral Electronc, vol. 55, no. 12, Dec [26] R.F. Waller, C.F.Landy, "Aement of nterbar current n double cage nducton motor wth broken bar," IEEE Tran. Energy Conv., vol.9, n.1, pp , March [27] A.O. D Tommao, R. Mcel, G. Rcco Galluzzo, "Montorng and Dagno of Falure n Squrrel-Cage Inducton Motor Due to Cracked or Broken Bar", IEEE- SDEMPED 2011, Bologna, 2011, pp [28] Y. Grtl, A.O. D Tommao, F. Flppett, R. Mcel, C. Ro, "Invetgaton of Motor Current Sgnature and Vbraton Analy for Dagnong Rotor Broken Bar n Double Cage Inducton Motor", Sympoum on Power Electronc, Electrcal Drve, Automaton and Moton, SPEEDAM 2012, June 14-16, Pa (Italy), pp [29] J. J. Rangel-Magdaleno, R. J. Romero-Troncoo, R. A. Oorno-Ro, E. Cabal-Yepez, L. M. Contrera-Medna, "Novel Methodology for Onlne Half-Broken-Bar Detecton on Inducton Motor", IEEE Tran. on Int. and Mea., vol. 58, no. 5, May [30] A. Stefan, A. Yazd, C. Ro, F. Flppett, D. Caade, G.A. Capolno, "Doubly fed nducton machne dagno baed on gnature analy of rotor modulatng gnal," IEEE Tranacton on Indutry Applcaton, vol. 44, no. 6, pp , Nov./Dec [31] Y. Grtl, L. Zarr, C. Ro, F. Flppett, G. Capolno, and D. Caade, "Advanced dagno of electrcal fault n wound rotor nducton machne," IEEE Tran. on Indutral Electronc (Accepted for publcaton), [32] A. Belln, F. Flppett, G. Francechn, and C. Taon, "Cloed-loop control mpact on the dagno of nducton motor fault," IEEE Tran. on Indutry Applcaton, vol. 36, no. 5, pp , Sep./Oct [33] E. Serna, and J. M. Paca, "Detecton of rotor fault n feld orented controlled nducton machne", IEEE IAS Annual Meetng, Tampa, FL, USA, pp. 1-7, October 8-12, [34] S. M. A. Cruz and A. J. M. Cardoo, "Dagno of rotor fault n cloed-loop nducton motor drve," IEEE IAS Annu. Meetng, Tampa, FL, 2006, pp [35] S.M A. Cruz, A. J. M. Cardoo, Analy and dagno of rotor fault n hgh-performance nducton motor drve, Proc. Int. Conf. Elect. Mach., Greece, 2006, pp [36] S. M. A. Cruz, A. Stefan, F. Flppett,A. J. M. Cardoo, "A New Model-Baed Technque for the Dagno of Rotor Fault n RFOC Inducton Motor Drve", IEEE Tran. on Indutral Electronc, vol. 55, no. 12, Dec. 2008, pp