Int ern at i onal  Journ al of  P ower E le ctr on i cs a n Drive  S ystem   (I J PE D S )   Vo l.   11 ,  No.   3 Septem be r   2020 , pp.  1333 ~ 1434   IS S N:  20 88 - 8694 DOI:  10 .11 591/ ij peds . v 1 1 .i 3 . pp 1333 - 1 34 3           1333       Journ al h om e page http: // ij pe ds .i aescore.c om   Model  refer ence  self - t uni ng fracti onal o rder  PID  contr ol b ased  on for a  powe r system  stabiliz er       M.   A. A bdel  Ghany 1 ,   Moh amed  A .  Sha msel din 2   1   Depa rtment of  El e ct ri ca l   Eng in ee ring ,   Fa cul ty   o Engi n ee ring   O ct ober   Univ ersit y,   Egypt.   2   Facul ty   of Engi nee rin g ,   Futur e Universit in  Eg ypt,   Egypt         Art ic le  In f o     ABSTR A CT    Art ic le  history:   Re cei ved   J ul   6 ,  201 9   Re vised  Feb 2 ,  20 20   Accepte Apr   8 , 2 0 20       Thi pape pr ese nts  novel   appr oac of  self - tuning  for  Modified  Frac ti on al  Order  PID   (MF OP ID)  depe nds  on  the  Model   R efe ren ce  Adapt i ve  Sys te (MR AS ).   The   p roposed  self - tun ing  cont ro ll er   is   appl i ed  to  Po wer  Sys te Stabi lizer   (PS S).  T aka j i - Sugeno   (TS)  fuz zy   lo gic   t ec hniqu i used  to   construc t   the  M FO PID   cont roll e r.   Th ob jecti v e   of  MRA is  to   update  the  five   p ara m eters  of  Ta k aj i - Sugen Modifie FO PI (TSMF OP I D cont ro ll er   onli ne .   For  diff e ren op era t ing  p oint of  PS S,  MRA is  appl ie d   t inve stig ate   the   eff ec t ive n e ss   of  proposed  cont rol le rs.  The   har mony  opti mization  te chn ique  used   t obt ai n   the  op t im al  p ara m eters  of   TSMF OP ID   con trol l ers  and  MRA par am e te rs.   For  diffe ren oper ating  point wi t diffe r ent  disturba nc und er  pa ramet ers   v ari a ti ons  th si mul ation  r esult s   are  obt ai n ed.  Thi is  to  show   tha Se lf - Tuni n of  TSMF OP I base on  (MRA S)  have  bet t er  pe rform an ce   tha n   the fi x ed   par a me t ers  TS MO FOPID c ontr oll er .   Ke yw or d s :   Fr act io nal or de P ID   Harmo ny r e sea rch ( HS )   M odel   ref e ren c e ada ptive  con t ro ( M RA C)   Power syste sta bili zer ( P SS )   Takaji - s ug e no  fu zz y   This   is an  open   acc ess arti cl e   un der  the  CC  BY - SA   l ic ense .     Corres pond in Aut h or :   M . A.  Ab del Ghan y,    Dep a rteme nt   of Elec tric al  E nginee rin g,   Faculty  of E ngineerin g Octo be r 6  Un i ver sit y,   Emai l:  mghany 1988@ hotmai l.com       1.   INTROD U CTION     Gen e rato e xc it at ion   co ntr ol  sy ste ms   co nt ai A utomat ic   Vo lt a ge  Re gula tors  ( AV R )   for  vo lt age   regulat ion  a nd  co nventi on al   Power   S ys te m   Stabil iz ers   (C PSS)  for   da m ping  mec han ic al   mode   os ci ll at ion s.   The  c ha nges  i operati ng  c onditi ons  of  PS is  c halle nge  to  update  t he  c on t ro ll er   p a ra mete rs  [ 1] .   T he refor e ,   the  ne st ud ie seek  t de sign   ad va nced   c on t ro te ch niques,  w hich  c ontr ollers  ada pt   with  the   co nt inuous   chang e   in   ope r at ing   points   [ 2 - 4].   T he  c onve ntion al   P ID  c ontr oller  is   co m mon  us e   in   se ver al   of  e ngin eerin app li cat io ns .   D ue  t o   the   str uc ture  s impli ci ty   an ea sy  pa ra mete t un i ng,  i is  su it able   f or  a   certai ope rati ng  po i nt.  In  a dd it ion ,   it perfor mance   is  good   f or  li near  an simple   syst ems  [5 6].  Sti ll the   be ha vior  of  PID   con t ro l i s li nea an ca nnot deal  w it the  hig h dist urb a nce  and   high  nonlinearit i co m plica te sy ste ms [5 7 8].   The  c urre nt  resea rch   di rected  to  us the  Fr act io nal   Order   PID  ( FO P ID)  c on tr ol  w her it   presents     the  no nlinear  face  of  P ID  c on t ro [9 - 11 ].   In   FOPI c ontrolle r,   t wo   a ddit ion al   par a m et ers  (the   f rac ti on al   int egr al   a nd  de rivati ve  gain s will   be  s uppl ementar to   i ncr ease   the   fle xib il it a nd  re li abili ty  of  co ntr oller  [12 - 14] T herefo re,  t he  dyna mic  pe r forma nce  of   FO P ID  co ntr oller  is  e nhanced  co m pa red  t   the con ven ti onal  PID c on tr oller [1 5 - 17].   At  dif fe re nt  op erati ng   points  for  certai s yst em,  ada ptati on  onli ne  was  use sel f -   tu ning  us i ng   for  the  s ys te m.   I this  case,  t he  f uzzy  lo gic  cal c ulati on s   nee d   a   long  ti me  a nd   ad diti on   e ffo rts  by  tr a nd  e r ror  is  performe to   ob ta in   normali zi ng  gains   sel ect ion  [ 18].  S o,  this   stu dy   r esor t   to   the   M RA t sel f - tu ning     the  T SMFO PID  onli ne   w here  it   has  simpl str uctu re,   ea sy  to   impleme nt  a nd  fast  ca lc ulati on s   [ 19 20].     Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2088 - 8 694   In t J   P ow  Ele Dr i   S ys t ,   V ol 1 1 , N o.   3 Se ptembe r   2020   :    13 3 3     1 343   1334   The  m odel - refe ren ce  ad a ptiv e sy ste m  (MR AS )  prese nts one  of  t he best a dap ti ve  contr ol  techn i q ue s. It   may be   reg a rd e as  a ada ptive   ser vo  s ys te m   in   w hich   the   desi re perf or m ance   is  e xpresse i te r ms   of  a   re fer e nc e   model,  w hich  giv es  the  desir ed  res pons t comma nd  s ign al   [ 20].   It  f or ces  t he  over al sy ste to  f ollow    the  beh a vior  of  pr esel ect e model  r efe rence T he  presel ect ed  m odel   c an  be  first   or  seco nd  or der  s ys te m   accor ding t th e point  of v ie w  the  desig ner a nd compli cat ed  d e gr ee  of the  s ys te [19 ].     In   this  Stu dy,  Takaji -   S ug e no  F uzzy   T ype  (T S - F uzz T yp 3)   is  desig ned   by  21  r ule   tria ngula membe rs hip s In   t he  TS MFO PI D   desi gn,  th powe of   S   operat or   for  F O PI D   is  li ed  between  zer λ  f o r   integral   f racti on  orde r   an zer <   μ   f or  de rivati ve.  Let   the  FOPI D   ope r at ed  via  t he   Ni ntege To olbo with   internall unkn own   fi ve  par a mete rs  ( kp,  ki,   kd,   λ  a nd  μ)  be   na med   as   ( TB FO P ID).   Wh il e,  a   m od i fied   F OP I wh ic ha ext ern al ly  unkn own  five  pa ram et ers  ( kp,  ki,  kd,  λ  a nd   μ c on st ru ct e de s ign e by  T f uzzy   is   name a T S M F OPID   [ 21] .   T he  perf or ma nce  of  t he  T S M F OPID   base on  model  refe ren ce   a dap ti ve   s ys te m   con t ro ll er  ca be  imp r ov e usi ng   dif fer e nt  typ e of  opti mi zat ion   te ch nique  Ha rm ony  S earch  (HS)  [22 23] This  pa pe pr esents,  new   com bin at io betwee m od i fied  F OPID  c on t ro ll er  base on  TS  te c hniq ue   (TSMF OPID ) a nd  M odel  r e fe ren ce  as a t une to  d esi gn a  n e a dap ti vel y o utput fee dback  contr oller for   PSS.       2.   POWER  S YST EM ST ABILIZ ER MOD EL   sin gle  mac hi ne - in finite   bu sy ste w ho s li near iz ed  i nc reme ntal  mod el   con ta ini ng   t he  volt age   regulat or  a nd  excit er  ca be   de monstrate by  the   bl oc di agr am   as   sho w in   Fig u re  1.  The   par a me te rs  of    the  s ys te a re   gi ven  in   Ta ble  [1 24 ].   A   num ber  of  ca ses  are   done  t hat  c ov e dif fe ren operati ng   po i nts  (normal hea vy   an li ght  loa co ndit ion s )   an pa rameters   va riat ion  in   the   pr ese nce   of   sever e   disturba nce.   These   cases  a r a ppli ed  t t he  s ys te m   wit op ti mal  FOPID  a nd  sel f - tun i ng  by  M od el   Re fer e nce  c on t ro l.     The  c onsta nts  [K1,  …,   K6]   in   the  nor mal,  hea vy  a nd  li ght  loa ds   a r giv e i Ta bl 2.  The   pa ram et ers  of  the s ys te m a re  giv e in  Ta ble  [2 25].             Figure.  1. Line ari zed i ncr e me ntal mo del  of s yn c hro nous ma chine wit a n e xcite a nd stabi li zer.       Table  1.  T he  paramet ers  of t he  sy ste m   Para m eter   Valu e   Para m eter   Valu e   Ka   400   D   0   Ta   0 .05   Ef d m ax   7 .3   Tdo   5 .9   Ef d m in   - 7 .3   M   4 .74   u m ax   0 .12   Kf   0 .02 5   u m in   - 0 .12   Tf   1     0       Evaluation Warning : The document was created with Spire.PDF for Python.
In t J  P ow Elec   & Dri S ys t   IS S N:  20 88 - 8 694       Mo del refe renc e self - tun i ng fr actional  order  PID co ntr ol ba sed o   ( M.  A. A bd el   G hany )   1335   Tab le   2.   O per a ti ng  c onditi ons  for   K _1 to K _6   K   OP1 =[1.0 1 .0]   No rm al  load   OP2 =[1.3 0 .9]   Heavy  load   OP3 =[0.8 1 .2]   Ligh t load   K1   1 .07 5 3   0 .62 3 4   1 .40 7 6   K2   1 .25 8 1   1 .28 1 3   1 .19 8 4   K3   0 .30 7 1   0 .30 7 1   0 .30 7 1   K4   1 .71 3 1   1 .71 2 3   1 .64 6 1   K5   - 0 .04 7 6   - 0 .20 9 1   0 .07 4 2   K6   0. 4972   0 .45 6 5   0 .54 8 8       3.   CONTR OL T ECHNIQ UES   This  sect io s hows   the  pro pose c on tr ol  t echn i qu e s.  Th first  te c hn i que  is  t he  Modi fied  F OPID   con t ro ll er  base on  Harmo ny   search T he  s econd  te ch ni que  is  the  sel f - t un i ng   of  m od i fied  F OPID  c ontr oller   base d on opti m al   M RA S.     3.1.   The FOPI D C on t rol   The  To olbo FO P ID  (TBF OP I D is   usu al ly  use to   s imulat the   F OP I D   co ntr ol.   It  has   fi ve   par a mete rs  int ern al ly  sel ect e by  de sig ner  in  on cl os e bl ock   as  s how in  Fi gur e.2 .   T ma ke   adap t     the  F OPID   c ontrol  onli ne  a   T akaji -   Suge no  F uzz (ty pe - fu zz y)  te ch nique  is   de vel op e FOPI has   e xter nal   five  te r minals  par a mete rs  as  s how in  Fig ure   3 [ 21 26].             Figure  2. The   blo c k diag ram wit inte rn al   fi ve  unknow n para mete rs  kp, ki kd, λ   an d μ     Figure  3. The   blo c k diag ram wit the  ex te rnal   five  par a mete rs   t o be s uitable  fo r mo del r e fer e nc e self - tun in g       The  desig ste ps   of   T SMFO I   and   TS M F O can  be  s ummari zed  a f ollow s   [ 21] T he  first  ste p,  le the  in pu T f uzzy  mem bers hip   functi ons  f or   t he  f racti on al   order of  th integral  a nd  der i vative    a nd   μ )   va lues   are   sel e ct   to  be   21  t riangular   me mbe s hip s   functi on s The   un i ve rse  of  disco urs val ues  a re   e qu al ly   distrib uted  ove the   range  [0,  2]   a nd  ha ve  t he ir  mid dle  ve rtic es  placed  at   t he  points  { 0,  0.1,  0.2,  0.3 0.4 0.5,  0.6,  0.7,  0.8 0.9,  1,   1.1,  1.2,  1.3 1.4,  1.5 1.6,  1.7,  1.8,  1.9,  2}  .T he  me mb e rsh i sel ec te by  21  tria ngula r   membe s hip s   as  show i Fi g ure   4.   T he  bl ock  of  TS MFO or  T SMFO and  TS M F OPI 21  ru le s   re presents   in Figu re. 5.           Figure  4. I nput  mem ber s h ip  of the  v a riables  of   λ a nd μ   Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2088 - 8 694   In t J   P ow  Ele Dr i   S ys t ,   V ol 1 1 , N o.   3 Se ptembe r   2020   :    13 3 3     1 343   1336       Figure   5. TS MFOP ID   co ntr ol le with  21 me mb e rsh i ps       The   sec ond  ste p,  rec ognize   th TS - F uzz for mu la   f or  the   frac ti on al   orde rs   of   integ ral  a nd  de rivati ve   par a mete rs  (λ  and  μ as  s ho wn   i Fi gure.  6.   If   t he  in put  is  λ  the  blo c diag ram  repres ents   TS M F OI  wh il if   the in pu t i s  μ the  blo c k diag r am r e pr e sents  TSMFO D.           Figure   6. Bl oc diag ram  t he f racti on al   orde r s of the  integ ra l and de rivati ve .       The  final  outp uts  of   t he  fu z zy  s ys te ms   th at   inferre for   the  T SMFO D   or  TS MFO I mp le me nted   us in Ce ntr oi d f or  t he def uzzi ficat ion   meth od  [21] :      =   .    ;    =   .    ;       ( 1 )     Wh e re:   λi  ,µi  ϵ   {0 0.1,  0.2,  0.3,  0.4,  0.5,  0.6 0.7,  0.8 0.9 1,  1.1 1.2,  1.3,1. 4,  1.5,  1.6,  1.7,  1.8,  1. 9,   2}for 2 1 r ules   Wλi is the  w ei g ht  of λi,   Wµi is t he  w ei gh t   of µi,   Fλi is t he o utput o TB FOPI  whose λ  value i s λi   Fµi i s the  outp ut of TB FOPD  who se  µ  value  is µi  [12].   Evaluation Warning : The document was created with Spire.PDF for Python.
In t J  P ow Elec   & Dri S ys t   IS S N:  20 88 - 8 694       Mo del refe renc e self - tun i ng fr actional  order  PID co ntr ol ba sed o   ( M.  A. A bd el   G hany )   1337   3.2.   Ha rm on y se ar ch optimi za tio n t ec hnique   The  c halle ng e   po i nt  in  the  PID  an FO P I con t ro ll ers  a re  sel ect ing   the  a ppr opria te   par a mete rs  f or  a   certai c on t ro l le pla nt.  T he r are  seve ral  methods   to  fin the   par amet e rs  of  F OP I c on t ro ll er   f or  e xam ple,   try  a nd  er ror  a nd  Zie gler - Nicho ls  meth od  but,  m os of  the se  te ch niques  are  r ough  road s.  I t his  pa pe r,   th e   harmo ny  sea rc opti mi zat ion  te ch nique  wi ll   be  us e t ob ta in   the   op t imal   value of  F OPI c on t ro ll er   par a mete rs  acc ordin t the  object ive fu nction as s how i n ( 2 [ 22] .     = 1 ( 1 ) ( +  ) + ( )   ( 2 )     Wh e re     is  the  ste ady   sta te   e r ror,     is  the  ove rsho ot  of  s ys te res pons e   is  the  set tl ing   ti me   and     is  the  rise   ti me.  Also th is  ob je ct ive  functi on   is  a ble  t co mpr om ise   the  desi gn e re qu i reme nts  us i ng   the  wei gh ti ng  par a mete valu (β).  T he  pa ra mete is  set   la r ger   t han   0.7  t reduce  ov e s hoot  a nd   ste a dy  sta te   error.  If   t his  pa rameter  is  adj ust ing   small er  t han   0.7  the  rise  ti me  and   set tl ing   ti me  will   be  re du ce d.   Ha rm ony  search   ( HS)  w as  sug gested  by  Z ong  W oo  Geem  i 2001   [27 ] It  is  wel known   that  HS   is  a   phe nomen on - mimi ckin al gorith in sp ire by  the   imp r ov isa ti on  proc ess  of  mu sic ia ns   [28 ] T he   init ia popula ti on  of   Harmo ny  M e mory   ( H M )   is  ch os e rand oml y.  H M   c ons ist of  Ha rm ony  M em ory  S olu ti on  (HMS)   vect or s .   Table  s hows   the  obta ined   par a mete rs  of  TSMFO PID  c on t ro ll er   base on  har m ony  searc opti miza ti on   te chn iq ue.        Table  3.   T S M F O PID  pa ramet ers.   TSM F OPID   p ara m ete rs   Kp   Kd   vd   ki   vi   Para m eters  valu es   9 .56 0 3   5 .35 0 6   0 .23 7 1 4   2 .59 2 6   0 .92 2       Both   of  c onve ntion al   to olbo of  FOPI D   an the   Ta kaji - Sugeno   (T S)  m od ifie FOPI ( TSMFO PID)   hav e   the   same   res pons e   th rough  t he  sim ul at ion   res ults  a differe nt  op e rati ng  co ndit ion s.   I a ddit ion,  it   is   pro vid e in   [ 21] .       = [             ( 1 , 1 )   ( 1 , 2 )   ( 1 , 3 )    ( 1 , 4 )    ( 1 , 5 )   ( 2 , 1 )   ( 22 )   ( 2 , 3 )    ( 2 , 4 )    ( 2 , 5 ) . . . . . . . . . . . . . . .   (  , 1 )   (  , 2 )   (  , 3 )      (  , 4 )      (  , 5 ) ]               ( 3     3.3.   The  sel f - tuni n TS MFO PID  ba se on   mo d el  reference  te ch nique   In  this  pa per ,   the  m odifie FO P ID  c on tr ol   par a mete rs  will   be  a dju st ed  on - li ne  us i ng  the   m od el   ref e ren ce   te ch nique.  The   Mo del  Re fe ren ce   Ad a ptive  C ontrol  ( M RAC)   c on si ders  high - eff ect ive ness  a dap ti ve   con t ro ll er  [19].  I t wor ks  as a n adaptive  ser vo   sy ste m   in  w h ic the  wan te pe rformance is  descr i bed  i f orm  of   ref e ren ce   m od el Fig ur e dem onstrat es  the  main  c onst ru ct io of   sel f - tu ning  m od i fied  F OPID  bas ed  on   model re fer e nc e tec hn i que. T he deta il s o f   Mod el  Re fer e nce  Adap ti ve  S ys t em d e rivati on  are   giv e i n [ 29] .           Figure  7. The   ov e rall  syst em  with self - tu ning T SMFO PID  bas e d on m od e l refe ren ce  techn i qu e     Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2088 - 8 694   In t J   P ow  Ele Dr i   S ys t ,   V ol 1 1 , N o.   3 Se ptembe r   2020   :    13 3 3     1 343   1338   The  tra nsfer  fu nction o f FO PID c on t ro l ca n be  descr i bed as  foll ow s   [ 29] .     ( ) ( ) = + 1 +   ( 4 )   =     ( 5 )     Assume t hat th e p la nt c an  b e   simpli fied   t a   first  order sy ste m as  obvi ou s   in the f ollow i ng e qu at io n.     ( ) ( ) =  + 1     ( 6 )     Wh e re          are   un known   par a me t ers.   Als o,  ass um e   that   the   model  ref e ren c ta kes   a   f orm   fir st  order s ys te m a s the  fo ll owin g rela ti on s hi p.     ( ) ( ) = + 1       ( 7 )     Wh e re         are  sel ect ed  by  desi gner .   Fr om e quat ions  [4 - 6 ]   can  con cl ud e t hat    =  + 1 ( + 1 + ) ( )     ( 8 )             = + 1 +   + 1 + 1 +   + 1     ( 1 + + . 1 +   + 1 ) = + . 1 +   + 1   (  + 1 + + . 1 +   + 1 ) = + . 1 +   + 1   = + . 1 +   + 1 + + . 1 +                                                                                                                     ( 9 )   =     ( 10 )   = [ + . 1 +   + 1 + + . 1 +  + 1 ]     ( 1 1 )   = [  + +  + . 1 + 1 ( + . 1 +  ) (  + +  + . 1 + 1 ) 2 ]     ( 1 2 )     The  ( 1 2 )   ca n b e re wr it te n     = [ (  + + . 1 +  + 1 . 1  ) (  + + . 1 +  + 1 ) 2 ]       ( 13 )   = [ (  + 1 ) (  + + . 1 +  + 1 ) 2 ]       ( 1 4 )     = [ (  + 1 ) (  + + . 1 +  + 1 ) ( + . 1 +  ) ]     ( 1 5 )     Fr om  ( 4 a nd ( 6 )     = [ 2 (  + + . 1 +  + 1 ) ]     ( 1 6 )     To  ac hi eve  the  desire d per for mance,  the  foll ow i ng con diti on  m us t   be h old.      + + . 1 +  + 1 = + 1     ( 1 7   Evaluation Warning : The document was created with Spire.PDF for Python.
In t J  P ow Elec   & Dri S ys t   IS S N:  20 88 - 8 694       Mo del refe renc e self - tun i ng fr actional  order  PID co ntr ol ba sed o   ( M.  A. A bd el   G hany )   1339   = 2 + 1       ( 1 8 )       Fr om t he MI T   ru le  ca n o btain  the  fo ll owin g rela ti on s hip      = . . 2 + 1       ( 1 9 )    = 1 . . + 1     ( 21 )   1 = . 2   ( 2 1 )   )  =     + ( 0 )       ( 2 2 )     Wh e re  ( 0 )   is t he  i niti al  v al ue of   pro portion al   ga in  , By  t he  sam e steps.     )  =     + ( 0 )       ( 2 3 )     Wh e re  ( 0 )   is t he  i niti al  v a lue  of  pro portion al   ga in  .     )  =     + ( 0 )     ( 2 4 )     Wh e re  ( 0 )   is t he  i niti al  v al ue of   der i vative  gain   .      = ln   ( ) [ ( + 1 + ) (  + + 1 + + 1 ) 2 1 (  + + 1 + + 1 ) ]     ( 2 5    = ln   ( ) [ + 1 +  1  1 (  + + 1 +  + 1 ) 2 ]     ( 2 6 )    = ln   ( ) [ (  + 1 ) (  + + 1 +  + 1 ) 2 ]     ( 2 7 )    = ln   ( ) [ (  + 1 ) (  + + 1 +  + 1 ) ( + 1 +  ) ]     ( 2 8 )     Also ,   f rom  ( 4 ) a nd ( 6 )      = 2 ln   ( ) [ (  + + 1 +  + 1 ) ]       ( 2 9 )    = 2 ln   ( ) . + 1     ( 30 )     = . . 2 ln   ( ) . + 1     ( 31 )     = 4 . . + 1                                                                                                                                           (3 2 )   4 = . 2 ( 0 )   ln   ( ) ( 0 ) = 2 . ( 0 ) . ln   ( )       ( 3 3 )   )  =      + ( 0 )     ( 3 4 )    = [  . . ln   ( )  + + 1 +  + 1  . . ln   ( ) ( + 1 +  ) (  + + 1 +  + 1 ) 2 ]    = [  . . ln   ( ) (  + + 1 +  + 1 1  ) (  + + 1 +  + 1 ) 2 ]       ( 3 5 )    = [  . . ln   ( ) (  + 1 ) (  + + 1 +  + 1 ) 2 ]   ( 3 6 )     Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2088 - 8 694   In t J   P ow  Ele Dr i   S ys t ,   V ol 1 1 , N o.   3 Se ptembe r   2020   :    13 3 3     1 343   1340    = [  . . ln   ( ) (  + 1 ) (  + + 1 +  + 1 ) ( + 1 +  ) ]     ( 3 7 )     Also ,  from  ( 4 ) a nd ( 6 )      = [ 2 . . ln   ( ) . (  + + 1 +  + 1 ) ]     ( 3 8 )    = 2 . . ln   ( ) . + 1       ( 3 9 )     = . . 2 . . ln   ( ) . + 1       ( 40 )    = 5 . . + 1                                                                                                                                          (4 1 )   5 = . 2 . ( 0 ) . ( 0 ) . ln ( ) = 3 . ( 0 ) . ln   ( )       ( 42 )   )  =      + ( 0 )     ( 43 )       4.   SIMULATI O N RESULTS   This  sect ion   de monstrate th simulat ion   r esults  of   fixe structu re  TS M F OPID  a nd  sel f - tu ning   TSMF O PID  ba sed  on  M RA a pp li ed   to   P SS  with  dif fere nt  op e rati ng  po i nts  (h ea vy  and  li ght  pa ra mete rs)   thr ough se ver a l t yp es  of  distu rb a nces.     Ca se  1: Fi xed s tructu re T SMF OP I D pe rform ance at  dif fer e nt ope rati ng cond it io n.     The   mec han ic a tor que  Tm   a nd  V ref  inc reas es  s udde nly  with  ste val ue  5%   in   case   hea vy,   li ght  a nd  normal   pa ram et ers  values The   res ults  a r de m on st rated  i Fig ur e .   8.  It   is  cl ea r   tha t   fix ed   st ru ct ur e   TSMFO PID   re sp onse   ca nnot   adap t   the   c hanges   in   ope rati ng  c onditi ons.   S o,  the   sel f - tu ni ng  bec om es   es sentia l   to obtai n hi gh  performa nce t hro ugh seve ral  op e rati ng c ondi ti on s a nd d ist urba nces.           (a)   (b)     Figure  8. The  s ys te dyna mic   res pons e  w it ste c hange  0.05       Ca se 2 :   Co mpa rison  betwee n t he fixe d para mete rs  T SMF OP I a nd sel f - tun in g f or   TS M F OPID  at li ght l oa conditi on.   In  case   of   li ght  pa rameters   va lues,   the   mec han ic al   t orq ue  Tm  a nd  Vr e f   e xpos es   to   ste change   5%   high.  T he  res ul ts  a re  dem onstr at ed  in  Fig ur e . 9 It  is  note t ha sel f - tu ning  TSMFO PID  ( M RA S)   r esp on se  has  low fluct uations, small   ov e rs hoot a nd it  r eac to  r e f ere nce  po i nt in smal l t ime .       Evaluation Warning : The document was created with Spire.PDF for Python.
In t J  P ow Elec   & Dri S ys t   IS S N:  20 88 - 8 694       Mo del refe renc e self - tun i ng fr actional  order  PID co ntr ol ba sed o   ( M.  A. A bd el   G hany )   1341       (a)   (b)     Figure  9. The  s ys te dyna m ic  r es pons es  at li gh t c onditi on       Ca se  3 :   Co m par is on   betwe en  the  fi xe structu re  TS M F OPID  a nd   sel f - tu ning  TSMFO PID  ( M RA S)  performa nce at  h ea vy con diti on.     The  pro po se con t ro ll ers   we r inv est i gated  by  co mp a rin the  dy namic  re sp onses   of  the   PSS  at   ste disturba nce   5%   in   the   mech anical   to rque   ∆Tm  a nd  V r ef.   Fig ur e .   10   il lustrate the   r esults  of  this   c ase.  It  is   obvious  t hat  th dynamic  respon s of  sel f - tu ning  TS M F OPID  (M R AS)  ha good  perfor mance  c ompar ed  t fixe d param et ers  TS MFO PID  contr oller whe re it has  small e r ov e rs hoot a nd s mall  sett li ng ti me .           (a)   (b)     Figure  10. Syst em dy namic  re sp onses  w it r efere nce trac ki ng.       Ca se  4:   T he   performa nce   of  t he   fi xe st ru ct ur e   TS MF OP I D   a nd  sel f - tu ning   TS MFOP ID  durin the   par a mete rs va r ia ti on s a nd h ea vy loa c onditi on.   To  i nv e sti gate  the  r obus tne s of  pro posed   con t ro ll ers t he   inerti c oe ff ic ie nt  inc reased   to  become  M =1 .5  of  nor mal  value  an the  distu r ban ce   for  ∆T an ∆Vr e re presen te by  0.0 s te cha nge  f rom  zer to  s eco nds,   then dec reased   by   0.0 f rom  se co nds  t seco nd s   a nd  finall dec rease by  0.0 as   s how i Figure.   11.   Fi gure.   12   sho w the   syst em  r esp on ses   dr i ve by  sel f - tun i ng  TS MFO PID  ( M RAS a nd  fixe par a mete rs  TS M F OPID   co nt ro ll er  w hen.  I is  cl early  s een  that   the  sel f - tu ning  TS M F OPID   ( MR AS )   ov e rc om es   th ese  var ia ti ons   a nd  giv e   good   re spo ns e   wit a   s ma ll   set tl ing   ti me,   th us  in di cat ing     the  ef fecti ven e ss  of  the  sel f - tun in T S M F OP I ( M RAS ove wide   range  of  pa r amet er  var ia ti on  an change  of  op e r at ing  c onditi on s     Evaluation Warning : The document was created with Spire.PDF for Python.
                          IS S N :   2088 - 8 694   In t J   P ow  Ele Dr i   S ys t ,   V ol 1 1 , N o.   3 Se ptembe r   2020   :    13 3 3     1 343   1342       Figure  11. Vari abl e d ist urba nc e f or  ∆ Tm a nd  ∆Vr e f           (a)   (b)     Figure  12. Syst em dy namic  re sp onses  w it r efere nce trac ki ng       5.   CONCL US I O N   no vel  sche m of   sel f - t un i ng  Ta kaji - S ug e no   Mo dified  F racti on al   Order  PI D   (T SMFO PI D baas  on   the  M odel   Re fer e nce  a dap ti ve  s ys te ( MR AS )   ap plied   on  P ower  S ys te Stabil iz er  ( PSS ).   TS  Fu zz te chn iq ue   is  use t c onstr uct  (TSMF OPI D) .   T he   obje ct ive  of  M odel   R efere nce  A da ptive  Syst em  ( M RA S )   tun es  t he  five  par a mete rs  of  Takaji - S ug e no   Mod i fied  FOP ID   c ontr oller  onli ne.   Dif fer e nt   op e rati ng  po i nt f or  PSS  we re  im pl emented   to   in vestigat t he  r obus t ness  of  pro posed   c on tr ollers.   The   ha rm ony  opti miza ti on   te chn iq ue   use to   obta in   the   op ti mal  pa ram et ers  of  propos ed  c ontr ollers.   The   simulat io res ults  pro vide   that   Self - T unin T SM F OPI bas ed  on  ( M RA S)  ha v e   bette pe rformance   tha the   fi xed  par a mete rs  TS M O FO P ID   Con tr oller.       REFERE NCE S   [1]   C.   Ch en,  “Coor dina t ed   Synthesis  of   Mult im a ch ine  Pow er  Sys t em  Stabilizer   U sing  An   Eff i cie nt  De ce n tra l ized  Modal  Contro ( DM C Algorit h m,   IE EE   Tr ans.   Powe r S yst. ,   vo l .   2 ,   no .   3 ,   pp .   54 3 550,   1987 .   [2]   A.  M.  A .   Gh any, “ Pow er  Sys te m A utom at i Vol tage  Regulator   De sign Ba sed  on  St at i Output   Feed bac PID   Us ing   Ite ra ti ve   Li n ea Matri Ine qu al i t y, ”  2008  12 th  In t.   Midd le   East  P ower  Sy st.  Con f.  MEP CON  2008 ,   vol .   1,   pp .   441 446,   2008 .   [3]   M.  J.  Prabu,   P.  Poongodi,  and  K.  Premkumar,   “Rot or  Pos it ion  Control   of  Brus hle ss   DC  Motor  using  Adapti v e   Neuro  Fuzzy   Inf ere nc Sys tem,   Middl e - East  J .   S ci .   R es. ,   vol .   24 ,   no.   7 ,   pp .   2395 2403,   2016 .   [4]   Y.  Ta ng ,   M.  Cui ,   C.   Hua ,   L .   Li,  and  Y.  Yang,   “Opti mum   Design  Of  Frac ti ona Order  PI   Λd  Μ  Control ler  for  AV R   Sys te Us ing  C haot i Ant   Sw arm ,   E xpe rt S yst.  Appl . ,   vo l. 39, n o.   8 ,   pp .   6887 6 896,   2012 .   Evaluation Warning : The document was created with Spire.PDF for Python.