Hybrid FLC/BFO Controller for Output Voltage Regulation of Zeta Converter

Transcription

1 JOURNA OF ENGINEERING REEARH AND TEHNOOGY, VOUME 4, IUE, JUNE, 07 Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter H. Elayd and M. Albakh Abstract Renewable energy sources are usually connected to the power grd va power converters. Zeta converters are very mportant for mcrogrd and smart grd applcatons. The objectve of ths paper s to desgn a Mamdan fuzzy logc controller (F) and a hybrd fuzzy logc controller wth the bacteral foragng optmzaton algorthm (F/BFO) to mprove and regulate the output voltage response aganst dsturbances lke the change n the voltage source or the load for the Zeta converter operatng n contnuous conducton mode (M). Analyss and comparson among smulatons of the open loop, closed loop fuzzy logc controller, and hybrd F/BFO controller results were performed for dfferent output voltages and for dfferent workng condtons such as the change n the voltage source or the load. The results show that there s a sgnfcant mprovement n the results for the proposed F/BFO controller. The desgns and smulatons were performed n MATAB/IMUINK envronments. The results were compared wth other results whch used the partcle swarm optmzaton (PO) algorthm. Index Terms Bacteral Foragng Optmzaton Algorthm, ontnuous onducton Mode, Fuzzy ogc ontroller, Renewable Energy ources, Zeta onverter. I INTRODUTION Power systems produce electrcty dependng on load demands. Over the years, load demands ncreased n devloped countres. Energy sources have lmtatons on relablty of the supply, and cause envronmental polluton, global warmng, and the rsk of occurrence of nuclear accdents; thus, a need for renewable energy sources was born [], such sources nclude wnd, solar, hydro, and geothermal []. Renewable energy sources are usually connected to power grds va power converters. The choce of the approprate topology for nverters depends on many factors such as the type of the renewable energy source and the total amount of power that wll be handled [3]. D to D Zeta converters are one type of converters that are used to nterface renewable energy sources to the grd. Moreover, Zeta converters are used n many applcatons lke supplyng sutable D voltage to modern portable electronc equpment whch are not drectly connected to the A mans, power qualty mprovements, power factor correcton, and ndustral applcatons. In ths paper, we wll use a Zeta converter for convertng and supplyng sutable D voltage from a D voltage source to a load. In June 00, Vuthchhay and Bunlaksananusorn used a lnearzed model Zeta converter n the M mode to regulate the output voltage aganst dsturbances [4]. In order for the output voltage to meet a desred value, a PWM feedback controller was used, then a PI controller was added to mprove the system response. In 0, Moaven, et.al. presened a model reference adaptve controller (MRA) wth back-propagaton neural networks (NN) to control the output voltage of the Zeta converter operatng n M [5]. In 0, Izadan, et.al. mplemented a model reference adaptve controller (MRA) to the Zeta converter operatng n M mode for output voltage trackng [6]. In 03, arkaw, et.al. studed the Zeta converter operatng n the M mode to regulate the output voltage usng a full-state feedback controller [7]. They presented the system model by the A technque. The small sgnal lnear model consdered two nputs to the system: the nput voltage and the load current. The feedback gan matrx K was found by two methods: the pole placement method and the lnear quadratc regulator (QR). They found that QR gave them better results than the pole placement method, because QR found the optmal control effort. But ther system model was complcated. In June 04, Ahmad and ultan studed the Zeta converter operatng n M mode to mprove ts output voltage, and to control the output voltage under dfferent workng condtons or dsturbances such as changes n the load resstance or nput voltage [8]. A fuzzy logc controller (F) and a F wth partcle swarm optmzaton whch s known as a hybrd F/PO controller were presented to acheve the control goal. They compared the results of the open loop system wth F and F/PO whch concluded that F/PO produced the best results. arkaw, et.al. work's s one of the few reported works n the lterature to present a hybrd F/PO controller, whch reduced the system modelng of Zeta converter for controllng the output 48

2 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) voltage. But the output response had small rpples. Ths paper presents a desgn of a fuzzy logc controller (F) to reduce the control complexty of the Zeta converter system n M mode to mprove ts performance under dfferent workng condtons such as the load and the voltage source dsturbances. The desgns and smulatons were performed n MATAB/IMUINK envronments. The man contrbuton s to use a new optmzaton algorthm whch s the Bacteral Foragng Optmzaton Algorthm (BFOA) to mprove the F performance by optmzng ts scalng gans, whch results n desgnng a hybrd F/BFO controller [9]. The effectveness of the BFO algorthm wll be proved va the mprovement of the F performance for dfferent workng condtons. Ths paper s organzed as follows: ecton presents Zeta converter and ts modelng usng the A technque. ecton 3 presents the fuzzy logc control desgn. ecton 4 presents the Bacteral Foragng Optmzaton Algorthm (BFOA) as an optmzaton method that wll be used to get the best performance for the F. ecton 5 presents the hybrd F/BFO controller desgn. ecton 6 presents the results and dscusson. ecton 7 concludes ths paper and presents the future work. ecton 8 presents the references. (a) M mode (b) DM mode Fgure Inductors currents waveforms n M and DM modes The modelng of the converter s represented as a state space model. As explaned n the next secton, the overall model s obtaned by the state space averagng technque (A) from two state space models by calculatng the weghted average of two sets of equatons usng the nomnal values of the tme spent n each crcut state as the weghts. A Descrpton of Each rcut tate When the MOFET swtch s ON, the dode s reverse based, thus - open crcuted as shown n Fgure 3 below. In ths state, the nductors and are n the chargng state, and the nductors currents are ncreasng lnearly. II ZETA ONVERTER AND IT MODEING Zeta converter s a 4 th order nonlnear D-D converter [0], as shown n Fgure, has two nductors each wth a D Resstance (DR), two capactors each wth an Equvalent eres Resstance (ER), and a dode. The Zeta Fgure 3 The equvalent Zeta converter crcut when the swtch s ON converter can operate n step up or step down modes to supply a load. The second state s when the MOFET swtch s OFF, the dode s forward based, thus - short crcuted as shown n Fgure 4 below. In ths state, the nductors are n the dschargng state, and the energes n and are dscharged to capactors and whch are the output parts respectvely, and the nductors currents are decreasng lnearly. Fgure Zeta converter crcut The nput to the Zeta converter s a D voltage. The Zeta converter crcut has an operatng swtch (MOFET). Zeta converters may operate n one of two operatng modes, the frst mode s the ontnuous urrent Mode (M), and the second mode s the Dscontnuous urrent Mode (DM). Fgure 4 The equvalent Zeta converter crcut when the swtch s OFF Wthn one swtchng perod T, M mode offers two crcut states whle DM mode offers three crcut states. Ths paper focuses on M mode. Fgure llustrates the dfference between M and DM modes n ON and OFF states []. To nsure that nductors currents are ncreasng and decreasng lnearly, the followng equatons must be satsfed [4]: 49

3 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) ( D) R r r ( D ) Df R R D ( DR ) r ( ) f R Where f s the swtchng frequency and D s the duty cycle of the swtch. B tate pace Modelng of Each rcut tate When the MOFET swtch s ON as shown n Fgure 3,the state space model s as follows: r d dt d r R R 0 r r ( ) dt r R r R dv v dt v 0 dv R dt ( r R) ( r R) v o r R R 0 0 r R r R v v When the MOFET swtch s OFF as shown n Fgure 4, the state space model s as follows: d ( r r ) 0 0 dt d r R R 0 r 0 dt r R ( r R) dv v dt v dv R dt 0 0 ( r R) ( r R) v s (0) v o r R R 0 0 r R r R v v tate pace Averagng Technque (A) Durng the frst state, the MOFET swtch s ON for an nterval DT, whle durng the second state the MOFET swtch s OFF for an nterval (-D)T. The averaged (overall) state space model for the Zeta converter s obtaned as follows [7]: A A D A ( D) av B B D B ( D) av D ( D) av In ths paper, we assume deal Zeta converter, where all D Resstances and Equvalent eres Resstances have a value of zero; thus, the state space model becomes as follows: d D dt D d D 0 0 dt D dv D D v 0 0 dt v 0 dv dt R v o v v The relaton between the nput and the output voltages n the deal Zeta converter s characterzed by the duty rato as follows [4]: D vo vs D For M mode, the crtcal values for the nductance and capactance n the deal Zeta converter are as follows [4]: () v s (4) (3) 50

4 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) ( D) R Df ( DR ) f D 8 f ( D) R 8 fr (5) Fgure 6 The change of the error MF The produced rpples n the nductors currents and n the capactors voltages n the deal Zeta converter are gven n terms of v s and f, the swtchng frequency, as follows [8]: Dv s f Dv s f Dv v s 8 f Dv v s 8 f III FUZZY OGI ONTROER (6) Fgure 7 The output voltage MF TABE The rule base of the F E / ΔE N N Z P P N N N N N Z N N N N Z P Z N N Z P P P N Z P P P P Z P P P P In ths paper, three lngustc varables are used whch are two nput varables to the fuzzy logc controller; (the error, and the change of the error), and one output varable; that s the control sgnal to the Zeta converter system after beng defuzzfed. Each nput varable has 5 trangular membershp functons; thus, formng 5*5 or 5 rules. The Mamdan nference system s used, and the centrod method s used as the defuzzfcaton method. The membershp functons and ther ranges for the three The Fuzzy Assocatve Memory (FAM) or the table of rules s shown n Table [8]: The membershp functons shown n Fgures 5-7 need to be tuned n addton to the F scalng gans for the nputs whch represent a PD controller, and for the output n order to get the desred output performance. Thus, the control s acheved by a F wth PD controllers. The tunng was performed manually as t wll be explaned n secton 5. IV BATERIA FORAGING OPTIMIZATION AGORITHM (BFOA) Bacteral foragng optmzaton algorthm (BFOA) proposed by Passno [] s a smulaton of the socal foragng behavor of Eschercha col bactera present n human ntestne. Fgure 5 The error MF lngustc varables are as shown n Fgures 5-7: Generally, ths type of bactera move for a longer dstance n a frendly envronment. The chemotaxs of bactera could be a contnuous swm, a swm followed by a tumble, a tumble followed by a tumble, a tumble followed by a swm, or a combna- 5

5 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) ton of them. [3]. Fgure 8 shows the swm and tumble modes Where J (, P( j, k, l)) s the ftness functon value to be added to the actual ftness functon whch s to be optmzed to present a tme varyng ftness functon, [,,..., ] T p s a pont n the p dmensonal search doman, and d attractant, w attractant, h, and repellant w are dfferent coeffcents repellant whch should be chosen properly. 3. Reproducton (a) Tumble mode (b) wm mode Fgure 8 Modes of an E.col bacterum for a bacterum. A Processes of BFOA If J ( ) s the problem to be optmzed, where s a p dmensonal vector, the four processes of the BFOA are as follows:. hemotaxs The chemotactc step s consdered to be a tumble followed by a tumble or a tumble followed by a swm. et P( j, k, l) { ( j, k, l),,..., } represent the poston of each bacterum of the populaton at the j -th chemotactc step, k -th reproducton step, and l -th elmnaton dspersal event, or smply. The poston of the bacterum n the next chemotactc step after a tumble can be represented as follows: ( j, k, l) ( j, k, l) ( ) () T ( ) ( ) Where s a vector n the random drecton whose elements le n [-,]. If the ftness values of the bacterum mproved after the tumble, t wll contnue swmmng untl the ftness value degrades, then t wll tumble.. warmng warmng means that the bactera send sgnals to each other to congregate nto hgh bacteral densty groups and move to reach the desred locaton. et j(, j, k, l ) represent the cost or ftness at the locaton of the -th bacterum ( jkl,, ). warmng can be represented as follows: (7) The reproducton means that the bactera whch have had suffcent nutrents wll reproduce an exact replca of tself, and the least healthy bactera wll de. The number of the reproduced bactera wll equal the number of the dead ones, thus, the populaton sze of the bactera wll be constant n the evoluton process. 4. Elmnaton and Dspersal Elmnaton and dspersal smulates the sudden envronmental changes or attacks that may occur n the real bactera, thus, a group of bactera may be klled, and others may move to some other places. Whle smulaton, ths reduces the trappng n a local optmal pont. V HYBRID F/BFO ONTROER In ths paper, we wll optmze the scalng gans for the normalzed manually tuned membershp functons by usng the ntegral of the absolute value of the error or IAE as the ftness functon. The scalng gans for the nputs and the output of the F wll be used as varables that wll be optmzed usng BFOA. In ths case, the controller s called hybrd F/BFO controller. The BFO algorthm wll produce tral solutons for the scalng gans, and t wll determne f they mnmze the error n the system response by usng the ntegral of the absolute value of the error as a ftness functon. Then, the best scalng gans J (, P( j, k, l)) J (, (,, )) j k l p [ dattractant exp( wattractant ( m m) )] m p [ h repellant exp( w repellant (( m m) )] m (8) Fgure 9 Hybrd F/BFO controller wll be selected for the best system response. Fgure 9 llustrates the process of the hybrd F/BFO controller to control a system plant. 5

6 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) TABE 3 rtcal values parameters VI REUT AND DIUION The smulaton of the open loop Zeta converter system, the desgnng of a fuzzy logc controller (F) for the closed loop Zeta converter system, and the desgnng of a hybrd F/BFO for the closed loop Zeta converter system were performed for dfferent output voltages 9,, and 5 V for the nomnal values under dfferent workng condtons such as load dsturbance, voltage source dsturbance, or both. omparsons were made between our results and the results of Ahmad, et.al [8] to demonstrate the effectveness of our results and our methodology. The desgns and smulatons were performed under MATAB/IMUINK envronment. A The Normal Open oop Zeta onverter ystem Analyss The averaged state space model n equaton (3) has sx varables that must be defned n order to fnd the state space matrces whch are,,,, R, and D. The crtcal values of,,, and n M mode manly depend on the swtchng frequency f, the oad R, and the duty rato D. The nductors currents and the capactors voltage rpples are also affected. The crtcal values or lmts and the nductors currents and the capactors voltage rpples are as shown n equatons (5) [4] and (6) [8] respectvely. The duty rato D can be obtaned as follows: Zeta converter parameters V O=9 V, D=0.48 V = V V O= V, D=0.5 V O=5 V, D=0.555 (mh) (mh) (µf) (µf) Table 3 llustrates the crtcal values or lmts for,,, and under dfferent D and dfferent V O when the swtchng frequency f = 5kHz and the load R 0. It s clear that we must choose values that satsfy all the crtcal lmts n order to desgn a Zeta converter system that s vald for convertng the nput voltage to the output voltages 9,, and 5 V, n whch we must choose mh, TABE 4 The Zeta converter parameters Zeta converter system parameters F R 5 khz 0 5 mh 5 mh 90 F 0 F V D O V O V V = V TABE The duty rato for dfferent voltages The output voltage (V O) V The duty rato (D) (9) 0.57 mh 3. F, and.5 F Table 4 llustrates the values for the Zeta converter system parameters that are used n ths paper. The normal open loop responses wth the reference voltages are shown n Fgure 0. The normal open loop systems performances are llustrated n Table 5. electng V V, then, for each V O, there s a duty rato D. Table llustrates the duty rato D for V O 9,, and 5 V. 53

7 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) Fgure 0 The normal open loop system responses TABE 5 The normal open loop Zeta converter systems performances V (V) D V O (V) The normal open loop Zeta converter systems performances O (%) t (ms) e (%) V O rpples (V) B Fuzzy ogc ontroller Analyss The smulaton of the three normal closed loop Zeta converter systems whch are for trackng the output voltages V O 9,, and 5 V when V V compared wth the three normal open loop Zeta converter systems for a smulaton tme t 0. sec s shown n Fgure. Fgure The response of the normal F closed loop and the normal open loop Zeta converter 54

8 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) The normal F closed loop systems performances are llustrated n Table 6. TABE 6 The normal F closed loop Zeta converter systems performance V (V) V Ref (V) The normal F closed loop Zeta converter systems performances O (%) t (ms) e (%) V O rpples (V) A comparson between Table 6 and Table 5 whch s for the normal open loop systems performances s shown n Fgure. The ystem Dsturbance Analyss Fgure The normal F closed loop Vs. normal open loop systems The smulaton of the open loop and the closed loop Zeta converter systems for the output voltages 9,, and 5 V wth a smulaton tme t 0.5 sec under fuzzy logc controller s performed here. hanges n the load current and the load voltage s followed effcently when the load changes, voltage source changes, or both occur; thus, protectng the load from damage or malfunctonng. 0 at tme t 0.05 sec to reach 40 at tme t 0. sec, then t wll change from 40 to reach 0 at tme t 0.5 sec. Fgure 3 shows the changes n the values of the gan /R. The response of the three closed loop. ystem Analyss wth the oad Dsturbance The load R s consdered to change lnearly sweepng the values , n whch t starts to change from Fgure 3 The load dsturbance n the /R gan sgnal Zeta converter systems compared wth the three open loop systems wth load dsturbance s shown n Fgure 4. 55

9 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) It s clearly shown that the open loop response s affected by the load change n whch the varatons n the open loop output voltage accordng to the load change ncrease as the converted output voltage ncreases, whle n the F closed loop systems, these varatons were mnmzed and are under control. Fgure 4 F closed loop and open loop systems response wth load dsturbance. The ystem Analyss wth the Voltage ource Dsturbance The voltage source V s consdered to change lnearly sweepng the values 3 4 V, n whch t starts to change from V at tme t 0.05 sec to reach V at tme t sec, then t wll change from V to reach 3 V at tme t 0. sec, then t wll change from 3 V to reach 4 V at tme t 0.5 sec, then t wll change from 4 V to reach V at tme t 0.5 sec. Fgure 5 shows the changes n the values of the voltage source. Fgure 5 The voltage source dsturbance sgnal Fgure 6 F closed loop and open loop systems responses wth voltage source dsturbance The response of the three closed-loop and open-loop Zeta converter systems wth voltage source dsturbance s shown n Fgure 6. 56

10 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) It s clearly shown that the open-loop response s greatly affected by the voltage source changes n terms of the large varatons n the open-loop output voltage. On the other hand, the F closed loop system handles and controls these varatons effcently..3 The ystem Analyss wth both the oad and the Voltage ource Dsturbances In the real mplementaton of the Zeta converter system, both the load and the voltage source dsturbances are expected to occur smultaneously, and ths s the worst case scenaro for the Zeta converter system. The response of the three open loop and the three closedloop Zeta converter systems for a smulaton tme t 0.5 sec wth both types of dsturbances s llustrated n Fgure 7. The dsturbances n ths case were changed lnearly n the same manner explaned prevously. Fgure 7 F closed loop and open loop systems responses wth both dsturbances Thus, we may conclude that the desgned fuzzy logc controller (F) handles and controls the response of the worst case of dsturbances effcently. TABE 7 The BFO algorthm parameters The BFOA parameters.4 HYBRID F/BFO ONTROER DEIGN In ths secton, V O s lmted to 5 V as the worst case scenaro for smulatng results. Furthermore, our results n ths secton wll be compared wth the results from the work of Ahmad, et.al [8]. The desgned hybrd F/BFO controller s smulated consderng both types of dsturbances are present, and n whch the F scalng gans wll be tuned usng the BFO algorthm. Table 7 llustrates the parameters used n the BFO algorthm. ymbol Value P 3 6 N 5 N 4 N re 4 N ed P ed 0.5 The BFO algorthm s mplemented by MATAB/IMUINK usng three MATAB m-fles: the frst m-fle s the BFOA man code; the second m-fle s a functon to run the Zeta converter system wth each bacterum whch s a tral soluton, that computes ts ftness or 57

11 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) performance usng the ftness functon whch s the ntegral of the absolute value of the error or IAE; and the thrd m-fle s the cell to cell attracton functon to smulate the swarmng behavor of the bactera n the populaton. The used ftness functon s as follows: 0.5 J IAE e dt (0) 0 The response of the closed loop Zeta converter system and the reference voltage V Ref =5 V for a smulaton tme t 0.5 sec wth both types of dsturbances, prevously appled, s llustrated n Fgure 8. Fgure 8 The closed loop Zeta converter response usng the F/BFO controller Table 8 llustrates the performance of the normal closed loop F/BFO Zeta converter system when V Ref =5 V compared wth the normal open loop and F closed loop systems. TABE 8 Zeta converter system performance comparson for dfferent system desgns, V Ref =5 V ystem desgn Zeta converter systems performance comparson for dfferent normal system desgns O (%) t (ms) e (%) V O rpples (V) Open loop system F closed loop system F/BFO closed loop system Frst f we compare our desgns wth one another, table 8 shows that the F/BFO controller handles and controls the load and the voltage source dsturbances more effcently than the other controller types developed n ths work. The performance of the F/BFO closed loop controller, when compared to the F closed loop controller, mproved the overshoot by %, the steady state error by 44%, and the output voltage rpples by 0.8%. However, there was a 9% ncrease n the settlng tme as a drect result of decreased overshoot, owng to the ncreased dampng ntroduced to the system, whch naturally ncreases the settlng tme. When comparng our results wth the results of Ahmad, et.al [8], we note that they made smulatons of Zeta converter system usng 9,, and 5 V as output voltages for the open loop system when the nput voltage V = V. They also used 9,, and 5 V as reference voltages for the F closed loop system, and they optmzed the scalng gans of the F usng Partcle warm Optmzaton Algorthm (PO) whch resulted n desgnng the hybrd F/PO controller. The Zeta converter crcut parameters used n [8] are llustrated n Table 9. 58

12 H. Elayd and M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter (07) TABE 9 The system parameters used n [8] The Zeta converter system parameters used n [8] f R 5 khz mh 0.5 mh 900 F 000 F As prevously dscussed, Table 3 llustrated the crtcal lmts of,,, and when f 5 KHz and R 0. Thus, from Table 9, we can conclude that the selected values for and n [8] dd not satsfy the crtcal lmts. These crtcal lmts guarantee that the currents n and are ncreasng and decreasng lnearly whch guarantee that the average current n the load R s equal to the average current n the output nductor. On the other hand, Table 0 compares our results wth Ahmad, et.al [8] results for the normal systems when V Ref = 5 V. TABE 0 ystem performance comparson between our results and Ahmad, et.al [8] for V Ref =5V Zeta converter systems performance comparson for dfferent normal system desgns ystem desgn O (%) Ahmad, et.al [8] t (ms) Our desgn e (%) O (%) t (ms) e (%) Open loop system F closed loop system F/PO closed loop system F/BFO closed loop system Thus, we can conclude that our work compared wth Ahmad, at.al [8] gave better results where all of the crtcal lmts of the Zeta converter system parameters were satsfed, the open loop system performance was very good n terms of the overshoot, settlng tme, and steady state error; our desgned F clearly mproved the open loop performance, and the BFOA for the hybrd F system gave better results than PO wth regards to the overshoot and the settlng tme. The steady state errors n our results n the F closed loop and the hybrd F systems were close to the steady state errors n [8]. VII ONUION AND FUTURE WORK In ths paper, the smulatons of the open loop Zeta converter system, the desgnng of a F and a hybrd F/BFO controllers were performed under MATAB/IMUINK envronment for the nomnal values and for dfferent workng condtons such as the load dsturbance, the voltage source dsturbance, or both for the dfferent output converted voltages 9,, and 5 V when the voltage source was V. better than the open loop performance, n whch t mproved the overshoot, the settlng tme, and the output voltage rpples wth a very small ncrease n the steady state error for the dfferent output converted voltages and the dfferent workng condtons. The hybrd F/BFO controller performed better than the F controller, n whch t added mprovements to the overshoot, the steady state error, and the output voltage rpples wth a very small ncrease n the settlng tme. A comparson between our results and Ahmad, et.al [8] results whch used the F/PO controller for the reference voltage 5 V was performed. The comparson led to conclude that our results were better n terms of the overshoot and the settlng tme for the open loop systems, the F closed loop systems, and the hybrd F controller. The steady state error n our results for the open loop systems was better than n [8], whle n the F closed loop and the hybrd F systems, the steady state error was close to the steady state error n [8]. Thus, we may conclude that BFOA s compettve n comparson wth the PO n whch t presented better and more compettve results n the hybrd F system. The F usng Mamdan nference system performed 59

13 Frst A. Author, econd B. Author., and Thrd. Author / Research Name (06) Future work may nclude: usng the BFO algorthm to optmze the rule base component or the membershp functons of the F, usng the mproved BFO (IBFO) algorthm, usng ugeno nference system n the F desgn, usng type- fuzzy logc system n the F desgn, or usng other optmzaton algorthms wth the F such as Genetc Algorthm (GA) or Ant olony Optmzaton (AO) algorthm. REFERENE [] R.. Vero and F.. dos Res, "Desgnng closed-loop controllers usng a Matlab dynamc model of the Zeta converter n DM," 0th IEEE/IA Internatonal onference on Industry Applcatons(INDUON), 0. [] A. Kumar, H. A. Gftson, V.A. Rnoj, G.A. Jebaman, R. Balakrshnan, and M.. hnnathampy, "olar Energy Implementaton wth Grd Interfacng." Internatonal Journal of Advanced Research n Management, Archtecture, Technology and Engneerng (IJARMATE), 05. (): p. 9-. [3] E. F. amacho, T. amad, M. Garca-anz, and I. Hskens, "ontrol for renewable energy and smart grds." The Impact of ontrol Technology, ontrol ystems ocety, 0: p [4] E. Vuthchhay and. Bunlaksananusorn "Modelng and control of a Zeta converter", IEEE Internatonal Power Electroncs onference (IPE), 00. [5] B. Moaven, H. Abdollahzadeh, and M. Mazooch, "Adjustable output voltage Zeta converter usng neural network adaptve model reference control," nd IEEE Internatonal onference on ontrol, Instrumentaton and Automaton (IIA), 0. [6] A. Izadan, P. Khayyer, and H. Yang, "Adaptve voltage trackng control of zeta buck-boost converters." IEEE Energy onverson ongress and Exposton (EE), 0. [7] H. arkaw, M.H. Jal, T. A. Izzuddn, and M. Dahar, "Dynamc model of Zeta converter wth full-state feedback controller mplementaton." Internatonal Journal of Research n Engneerng and Technology (IJRET), 03. (08): p [8] A. H. Ahmad and N.. ultan, "Desgn and Implementaton of ontrolled Zeta onverter Power upply." Amercan Journal of Electrcal and Electronc Engneerng, 04. (3): p. -8. [9] M. Albakh, Hybrd F/BFO ontroller for Output Voltage Regulaton of Zeta onverter, M Thess, Islamc Unversty of Gaza, 06. [0] H. ra-ramrez and R. lva-ortgoza, ontrol desgn technques n power electroncs devces. 006: prnger cence & Busness Meda. []. Manktala, wtchng power supples A to Z. 006: Elsever Inc. []. Das, A. Bswas,. Dasgupta, and A. Abraham, Foundatons of omputatonal Intellgence Volume 3: Global Optmzaton. Vol : prnger. [3] H. upryono, Novel bacteral foragng optmsaton algorthms wth applcaton to modellng and control of flexble manpulator systems. PhD Thess, The Unversty of heffeld, Unted Kngdom 0. Hatem A. Elayd receved a B.. degree n Electrcal Engneerng from olorado Techncal Unversty n 990, and M.. and Ph.D. degrees n Electrcal Engneerng from New Mexco tate Unversty n 99 and 997, respectvely. He s currently an assocate professor at the Electrcal Engneerng Department, the Islamc Unversty of Gaza. He held several poston such as department head, assstant dean, and head of the Resources Development enter, head of qualty assurance unt, and Assocate Vce Presdent for Academc Affars. Hs research nterest ncludes control systems wth concentraton on optmal control, robust systems, convex optmzaton; n addton to qualty assurance n hgher educaton and uversty governance. He conducted several studes and consultatons n Palestne and the regon. He s certfed as a regonal subject and nsttutonal revewer. He s a member of IEEE, IAM, Tau Alpha P, AM, Palestne Engneerng Assocaton, and Palestne Mathematc ocety. He served as edtor board member, member of techncal councl, member of scentfc commttees for several local, regonal and nternatonal journals and conferences. Mohammed alsbakh got the bachelor degree n Electrcal Engneerng from the Islamc Unversty of Gaza n 006. Then he got the Mc degree n the Electrcal Engneerng\ontrol ystems from the same unversty n 06. For more than 0 years, most of hs tranngs and work experences focuses n the feld of computng and nformaton technology n the varous felds whch nclude computer networkng nfrastructure, Mcrosoft systems engneerng, computers and computer networks mantenance n both of hardware and software felds, helpdesk, techncal support and offce applcatons. Mohammed alsbakh currently works n Mnstry of Health (MOH) as omputer Networks Engneer and Techncal upport. Also he works as techncal nstructor n the IT feld n the prvate sector. 60