Adaptive Repetitive Control of a PWM Inverter for AC Voltage Regulation with Low Harmonic Distortion

Shing-Chung Yeh and Ying-Yu Tzou, Member, IEEE

Power Electronics & Motion Control Lab., Institute of Control Engineering National Chiao Tung Univ., Taiwan, R.O.C.
Abstracf% adaptive repetitive control scheme is proposed and applied in control of a PWM inverter used in highperformance ac power supply. The proposed control scheme can adaptive eliminate periodic distortions resulted by unknown periodic load disturbances in an ac power supply. The proposed adaptive repetitive controller consists of a voltage regulator using state feedback control, a repetitive controller with tuning parameters, and an adaptive controller with recursive leastsquares estimator. This adaptive repetitive controller used for ac voltage regulation has been realized by using a single-chip digital signal processor @SP) TMS320C14 from Texas Instruments. Experimental verifications have been carried out on a 2 kVA PWM inverter. Simulation and experimental result show that the DSP-based adaptive repetitive controller can achieve both good dynamic response and low THD under large load disturbances and plant uncertainties.

I. INTRODUCTION
Closed-loop regulated pulsewidth modulated (PWM) inverters have been widely applied in various type of ac power conditioning systems, such as unintermptible power supply (UPS) systems, automatic voltage regulators (AVR), and programmable ac source (PAS). One major requirement of these applications is that the system is required to maintain a low-distortion waveform under transient or periodic load disturbances. Some research have been carried out on the closed-loop regulation of PWM inverters to achieve good dynamic response and most of them were focused on the transient response improvement by using instantaneous feedback control [ 1]-[7]. Although satisfactory results have been obtained for transient load disturbance, they leave periodic distortions in the output waveform when the load disturbance is periodic. Repetitive control theory [SI-[lo] originated from the internal model principle [ 1 I ] provides a solution to eliminate periodic error occurred in a dynamic system. A repetitive controller can be viewed as a periodic waveform generator augmented within the control loop of a control system, which is closed-loop regulated by a feedback controller, so that the periodic errors can be eliminated. A number of repetitive control schemes have been developed and applied to various industrial a p p k t i o n s [12]-[14]. Nishida and Haneyoshi [15], and Hane, um.et al. [ 161 had applied the repetitive control technique to eliminate periodic distortions resulted in a PWM inverter. In their approaches, the repetitive controller was designed based on the exact model of the closed-loop
0-7803-2730-6195 $4.00 0 1995 IEEE

controlled system with one-step prediction. Therefore, the performance and stability of such repetitive control systems is highly dependent upon the robustness of the original feedback control system. Torelax the stringent stability requirement of a repetitive control system, a modified repetitive control scheme with finite frequency mode elimination has been developed [ 171. However, the convergence of such system will be deteriorated due to plant uncertainties. For time-varying or system with large plant uncertainties, adaptive repetitive control schemes were developed to eliminate periodic errors [181-[ 191. Although these methods can track the changing plant dynamics, they have the drawback that the number of parameters to be estimated are proportional to the frequency modes selected to be canceled. In this paper, we have developed a new adaptive repetitive control scheme that employs an auxiliary compensator to stabilize the closed-loop system and its parameters are tuned by an adaptive tuning controller which recursively on-line identify the plant dynamics. The adaptive repetitive controller will guarantee the closed-loop stability under plant variations and, at the same time, eliminates periodical errors resulted by all frequency modes below the closed-loop bandwidth. This paper is organized as following. In Sec. 11, we firstly introduce the operational principle, and then we derive the stability criterion and convergence of the periodic error of a repetitive control system. Sec. I11 describes the proposed adaptive repetitive control scheme. Sec. IV describes the modeling and the DSP-based repetitive control of a PWM inverter for ac voltage regulation. Sec. V gives the simulation and experimental verification of the proposed control scheme and Sec. VI is the conclusion.

11. REPETITIVE CONTROL SYSTEM

A . Principle o f Repetitive Control A servomechanism system is required to regulate the controlled variables to reference commands without steadystate error against unknown and unmeasurable disturbance inputs. In servomechanism system design, the internal model principle proposed by Francis and Wonham [ 11] plays an important role. The internal model principle states that the controlled output tracks a set of reference inputs without steady-state error if the model which generates these references is included in the stable closed-loop system. The

157

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY TIRUCHIRAPALLI. Downloaded on May 23,2010 at 08:43:43 UTC from IEEE Xplore. Restrictions apply.

Repebtive

Tracking

I I

L3asic Servo Plant

(21) Repebtive

F'

transfer function of the plant in which it is closed-loop regulated by a tracking controller, S(z-') and are the auxiliary compensator of the repetitive controller, r(k) is the reference signal, v(k) is system output. e(k) is tracking error. and r,(k) is compensated reference command. The transfer function from the disturbance input d(k) to the tracking error e(k) in Fig 2 is

e(.-')

Controller

where E@') and D(z-')are z-transforms of e(k) and tl(k). If d(k) is a periodic disturbance with period of N , then the Fourier series representation of d(k) can be expressed as
N-l

,=O

Pig. 1. Conligurations ofrepztitivc control system

where {c,,} denotes the Fourier coefficients and its corresponding frequency response in s-domain is
I

..............

f f ( j m )= fI(z-'

)Iz

s.!llll

(31

I n a special case, if Q(z-')= 1 and P(z") is stable. we can find

l f I ( j o ) ) l = O at m = 2 n n / N , n = O , . . N - I . (4)

This reveals that these frequency modes of the periodic error have been eliniinated by the repetitive controller, thus perfect ................................ I tracking is achieved in this condition. However, such a perfect tracking imposes a stringent stability criterion in the Fig 2 Block diaqam o f a disci-etc.time repetitive control system. synthesis of S'(z-'), In practical situation, we can relax this stability criterion by choosing Q(z-') a low-pass filter or a internal model principle therefore reveals that the high constant less than unit such that accuracy asymptotic tracking properties for periodic exogenous inputs can be achieved by locating the model IH( j m ( < p ( j w ) at w = 2 n n l N , n = O,..N --I (5) which generates a set of periodic signals. Repetitive control is a control scheme, in which the frequency modes of the where p o w ) specifies the attenuation of the frequency modes periodic error to be eliminated arc included in the stable of the periodic disturbance control loops. Repetitivc control can be easily realized using B. Stability Awa&sis microprocessor-based digital controller and with the advance of high-performance microprocessor and digital signal From Fig. 2 . we can find processor, more frequency modes can be included in the control loops. This reveals the feasibilit! of implementation of superprecision servomechanism system. and In a repetitive control systcm. a rcpetitivc controller is insertcd in thc control loop i n addition to the conventional tracking controller. There are various control configurations for the repetitive control systems. Fig. 1 shows two basic Eliminating I ( z ') from (6) and (7), we can get control structures of the repetitive control system. Fig. l(a) E(ZL1=E(zI)z "(Q(z')-P(z~)S(z')) sho\vs a cascaded type repetitive controller, in which it is cascaded as an outer loop controller. Fig. l(b) provides a + ( I - P ( z '))(I - Q(z '12 N ) R ( Z ). (8) feedforward path to the repetitive controller. The major purpose of the tracking controller is to improve the system Fig 3 shows the block diagram representation of (8) and we transient response and make i t insensitive to esternal can observe that if disturbance. ul~ile the repetitive controller is to reduce IQ(z ' 1- P ( z ' ) S ( Z ' ~ i a r< l 1 f a all w (9) periodic error resulted by periodic reference or disturbances. Fig. 2 shows the block diagram of the proposed repetitive and P(z-') I S stable, then e ( k ) is bounded which means the control system. where P(z-') represents the closed-loop system is stable
I

)I

158
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY TIRUCHIRAPALLI. Downloaded on May 23,2010 at 08:43:43 UTC from IEEE Xplore. Restrictions apply.

. .

Fig. 3. Block diagram representationofthe error convergence.

p x p
...

The design of S(z") and Q(z-')is a compromise between the degree of relative stability and the convergence rate of the Fig. 4. The proposed adaptive repetitive control scheme. periodic error elimination. For simplicity, if we choose Q(z-') to be a constant less than and close to unit. we can arbitrarily the stability where the parameters a,@) and a,@) are left to be identified. vev gain to choose an , ~ ( ~ - 1 ) criterion of (9). However. the periodic error may be still large. There are a number of well-known parameter estimation To satis@ both requirements (5) and (9), we can choose Q(z-') techniques that have been successfully applied to the to be a constant less than and close to unit and S(Z-') to be a identification problem [20]. The recursive-least squares digital filter with phase-lead characteristics, An optimal estimator (LSE) has the advantages Of unbias, fast choice of S(z-') in terms of large relative stability and fast convergence. and its aPPlicabilif3' to a wide range of l ) ~ ( ~ - ' )possesses application in which other statisticabestimation theories Inay convergence rate can be achieved when s(z a nearly zero-phase-shift characteristics [ 191. This can be be difficult to apply [211. In the aPPl*cation of an adaptive ) the inverse of ~ ( ~ - 1and ) control system, the most important consideration factor is its accomplished by choositlg ~ ( ~ - 1as augmented with a realizable low-pass filter with appropriate feasibility to be realized with an acceptable sanding *ate. cut-off frequency. The recursive LSE (RLSE) parameter identification algorithm used in this paper is: C. Convergence .4 nalysi.v P ( k - 2)&k - 1) 8 ( k )= &k - 1) + u + 4 ( k -I)'P(k--2)4(k-l)' The ie(z~')-S(z-')P(z-')l in the stability criterion of (9) can (13) ( y ,k ) - rc(k- 2) - 4 ( k - I)'&$ - 1)) also be viewed as a performance index for the convergence of periodic error. A smaller IQ(z-')-S(z~')P(--')( results a faster error convergence. The convergence index is defined as
h = (Q(z-') -P(z-')s(z-') *~ _ P(k - 1) = - P(k - 2) a l[

P(k - 2)qi(k - l)#(k - l)'P(k - 2) (14) a + $(k - 1)'P(k - 2)+(k - 1)

(13
I

1.

(10)
B ( k - l ) = [ a , ( k ) a,(k)l',

If h=O, the periodic error can be eliminated after one cycle. However, such a condition requires a perfect match of the plant model P(2.l) and this is obviouslv unrealistic. To avoid such situation, we define

@(k

- 1)

[-y(k - 1)

-y(k -2 ) ] ,

(16)

g Q@'> S(z-1) = ST

+ 1 P(z-')

(''I

where is a 'Onstant between zero and unit' and realization time constant
111. ADflrrvE REPETITIVE CONTROL SYSTEM

' is a

Fig. 4 shows the proposed adaptive repetitive control scheme. In addition to the conventional repetitive controller, an adaptive parameter tuner is included in the control loop to adjust the parameters Of Ls(z-') according to the identified plant Let ' ( k be the transfer function Of a second-order system and represented by
P ( k , z ~ '= )
Z
~2

1 + a , ( k ) z - '+a,(k)z-z

(12)

where &0) is the initial guess of the parameters to be identified, P(k) is a positive definite measure of the estimation error and its elements tend to decrease as the identification reaches its steady state, and CL is a forgetting factor to weight new data more heavily than old data. When tx=l, all data are weighted equally. For O<cc<l, more weight is placed on recent data than on older data. A snialler (x will result a fast convergence while it is also more sensitive to measurement noises. The PWM inverter used in an ac voltage regulator is usually connected to a bridge-rectifier RC load with its output waveforms as shown in Fig. 5(a). Fig. 5 ( b ) shows the estimated parameters of (12) RLSE, in -,hich, we see that rapid convergence with large oscillation has occurred whenever the rectifier switches change their state. One way to avoid this phcnomena is that when the load transition i s detected, a set of nominal parameters are used to set the RLSE parameters. Fig. 5(c) shows the estimated parameters of (1 2 ) using the modified RLSE.

159
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY TIRUCHIRAPALLI. Downloaded on May 23,2010 at 08:43:43 UTC from IEEE Xplore. Restrictions apply.

PVIMINMRTER

LCFILTER

LOA0

/SENSOR FILTER

Fig. 7. Hardware configuration ofthe digital controlled PWM inverter

Fig. 5. Parameter identification of a PWM inverter connected with a bridgerectifier RC load, (a) output voltage and current waveforms, @) estimated parameter usmg the RLSE, (c) estimated paramcter using the modified RLSE,(d) estimated parameters ofthe averaging model using modified RLSE.
vonags sensor and filter

Fig. 8. State feedback control block diagram of PWlM inverter with nominal load R = 40 ohm.

and
Fig. 6. Diserete adaptively repetitive controller.

Fig. 5(d) shows the estimated parameters of the averaging model using modified RLSE method. The estimated Since there are switching ripples in the capacitor voltage parameters of the averaging model are derived using the and inductor current, they are sensed through low-pass filters. following algorithm Considering the dynamics in these filters, we can get

where N is the number of samples during one half period of the output waveform. The auxiliary compensator S(z-l)of the From (19)-(22), the state equation and output equation of repetitive controller is adjusted by rising the estimated the plant is parameters of Z 2 ( n ) as shown in Fig. 6. i ( t ) = k ( t ) + Bu(t) , (23)

a,(??).

IV. MODELING AND CONTROL OF PWM INVERTER SYSTEM

The hardware configuration of the proposed DSP-based digital controlled PWM inverter system is shown in Fig. 7, in which, the combination of H-bridge PWM inverter, LC filter, and Rectlfier-type RC load is considered as plant. A. Plant Modeling The capacitor voltage vc and the inductor current iL are chosen as the state variables and the system dynamic equations can be derived as
160
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY TIRUCHIRAPALLI. Downloaded on May 23,2010 at 08:43:43 UTC from IEEE Xplore. Restrictions apply.

u ( t ) = v, ( t ) ,

(27)

A=

-20
-40 1
deoree
5

B =[1/L 0 0 01,

and
R
-400 I
10

. . . . . .
2

I
lo3
rad/sec

...... I
10
5

lo4

R+r,

R+r,
2 ,
1.5 1

(a)
I

The corresponding discrete-time model can be derived as x(k+l) = G p x ( k ) + H p u ( k ) ,

(3 1)

1.

AQ(2-l) =

e(.-')-

O.SP,(Z-~)P,~~(Z-')

0.5

G , =eAT,

(33)
(34)

E o
-0.5

HP = A-'(eAT - I)B,

-1

-1 5

1
-2
15
1

stable boundary

0.5

0 5

1 5

real

where Tis the sampling period.

(b)
Fig. 10. (a) Frequency responses of the state feedback controlled PWM inverter
system P,(i ) and the approximate model P,(z'). (b) Nyquist plot of Q(z')1

B. State Feedback Control

Considering the state feedback control block diagram in Fig. 8. The control law can be derived as
u(k)= k,u,(k)-k,u,(k)-k,v,(k),

0.5Pn(z1~,(zL which ), shows stability of the repetitive control system when

S(Z-I)=llP*(i')

(37)

where uref(k)is a table of reference command stored in (43) memory of DSP, the state feedback gains k,, , k, , and k, can be determined by method in [221. Combining (31), (32), and It should be noted that the plant model P,(z-') is derived based (37), we can obtain the state space equation of the digital on a nominal operating condition and in practical condition it controlled system as will encounter large model uncertainties due to large load variations.
C. Design Example

Table I lists some of the key parameters of the constructed PWM inverter system for 60 Hz, llOV (RMS) ac voltage regulation. A a single-chip DSP TMS320C14 from Texas Instruments has been adopted to realize the proposed adaptive repetitive controller. The sampling frequency of the digital controlled is 15 kHz and there are 250 samples in each cycle and of the sinusoidal output. The sampling rate of the adaptive K = [0 0 k, k2]. (42) parameter tuner is 120 Hz and it adjust the control parameters of the repetitive control every half cycle. The discrete-time transfer function from reference command The state feedback gain k, and k, are determined to to output voltage is minimize the output voltage distortion due to transient load where
161
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY TIRUCHIRAPALLI. Downloaded on May 23,2010 at 08:43:43 UTC from IEEE Xplore. Restrictions apply.

TABLEI. PARAMBTERS OF THE PWM JNVERTER SYSTEM

Item

I Symbol INominal value I

f,

Unit

Sampling Rate Filter Inductor Filter Capacitor Inductor ESR Cap serial resistor Noininal load
DC link voltage
Output voltage

0.6 33
(I 5

r
r L

PF

n
R
R
Volts
Volts (rms)
0.5
0.4

rc

10

,
~

c
' j
0

40 300
110

samples
(8)

'
U -

0.3

0.2
0.1

disturbance and the feedforward gain k,, is a scaling parameters to let the system have an unit gain at 60 Hz. With the given parameters as shown in Table I, the kl = -0.9 and k,=-0.45 are selected as state-feedback gains and the corresponding feedforward gain k, is 0.56. The nominal transfer function Pn(z-') of the digital state-feedback controlled PWM inverter is
P"(Z

0
-0.1

-0.2

-0.3
-0.4

-1

I '
50 60

-0.5

10

20

30 cycles

40

(b')
Fig. 10. (a) Simulation results ofthe m o r convergence using the proposed adaptive repetitive control scheme, (b) Time responses of the estimated parameters ofthe approximate average model.

0 53 1Oz ' + 0 3h55z - 0 04852 + 0 O O O ~ Z - ~ ' ) = 1 - 0 5 1 8 2 ~ '+04789z '-0127iz ' - 0 0 0 1 5 ~ - ~ (45)

The recursive LSE method has been used to identifjr the transfer function of the closed-loop controlled PWM inverter based on a second-order approximate model of (12). The identified plant model Po(z-')is

factor of 3. Fig. I I(a) shows the time responses of the output voltage and current of the PWM inverter using the digital state feedback control and Fig. Il(b) shows the same condition when the adaptive repetitive control is included. Fig. 12 shows the error convergence of the PWM inverter 2 Z 1; ( z - ') = __-__ (46) under repetitive control and it can be observed that it elapses 1-0 3 1 1 '~ + O 2962 about 12 cycles for the settling of the periodic error. For ;I 60 Hz output, it corresponds to 0.2 sec to elirniriate the periodic The frequency responses of P,(z-') and P,(z-') are shown i n disturbance resulted by a step changed rectifier-RC load. Fig. 9(a) and it can be observed that there are close resemblance between P,(z") and P,(z-'). The Nyquist plot of ~(z~')-P,(z~')Un(z~') is shown in Fig. 9(b) and it can be VI. CONCI,IJSION observed that it is within the stabilit) boundary and this An adaptive repetitive control is proposed and successfully guarantee the stability of the repetitive control system. applied for the closed-loop regulation of a PWM inverter used

v. SIMULATION AND EXPERIMENTAL VERIFICATION

Fig. 10 shows the simulation results of the adaptive repetitive controlled PWM inverter for ac voltage regulation. Fig. l0(a) shows a 3-dimension plot of the error convergence of the PWM inverter connected with bridge-rectifier RC load, Fig. 1001) shows the time responses of the estimated parameter estimation of a second-order approximate average model. The experimental verification of the proposed adaptive repetitive control scheme is carried out on a 2 kVA PWM inverter connected with a rectifier-RC load with current crest-

in high-performance ac power supply. Simulation and experimental results show that the proposed control scheme can effectively eliminate periodic waveform distortion resulted by unknown periodic disturbance. Compared with the conventional repetitive control methods, thc proposed adaptive repetitive control scheme can not only achieve faster convergence rate. it also guarantees stability robustness under large load variations. The THD for the rectifier-RC load of current crest factor 3 is 8% by using the state feedback control and it can be reduced to about 1% within 0.2 sec for 60 Hz rated output. An important merit of the proposed adaptive repetithe control scheme is that rt can be designed and implemented independently without knowing the exact

162
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY TIRUCHIRAPALLI. Downloaded on May 23,2010 at 08:43:43 UTC from IEEE Xplore. Restrictions apply.

Volt
200

Amp

150
100

50

-50
-100
-150

10

15

Time (msec)

20

Volt
200
150

(a)

100

50

0
-50
-100 -1 50
-200

.1
0

---- .~~ . . . . . . . . . . . i-.... . . . . . . ~.-------~-... L.... ----.-.------.... A . . . . . . . . . . . . . . . . . . . . . j -20 5 10 15 20

Time (msec)

(b)
Fig. 1 1. (a) The 60 Hz output waveforms for rectify load, Lvrrent crest fador 3 with only state feedback control, (b) output waveform when the adaptive repetitive control is applied.

Volt5

- - - - - - 10 -----I

-1

-I___

02

0' 4
Tiiriiilsarl

0 6

08

Fig. 12. Error convergence ofthe PWM inverter under repetitive control.

[4] A. Kawamura, T. Haneyoshi, and R. G. HoR "Deadbeat control PWM inverter with parameter estimation using only voltage sensor," IEEE Power Electron. Specialists' Con$, pp. 576-583, 1986. [ 5 ] C. Hua and R. G. HoR "High performance deadbeat controlled PWM inverter using a current source compensator for nonlinear loads," IEEE Power Electron. Specialists Con$, pp. 443-450, 1992. Tzou, "Discrete feedfonvard sliding mode control of a [6] S. L. Jung and Y. Y. PWM inverier for sinusoidal output waveform synthesis," IEEE fower Electron. Specialists' Con$, pp. 552-559, 1994. [7] Y. Y. Tzou I,. H. Ho, and R. S . Ou, "Fuzzy control of a closed-loop regulated PWM inverter under large load variations," IEEE fnd. E1ectronic.r Con$, vol. 1, pp, 267-272, 1993. [XI T. Inoue, M. Nakano, and S. Iwai, "High accuracy control of servomechanism for repeated contouring," in Proc. IOth Annual Symp. Incremental Motion Contr. Syst. andDevices, pp. 258-292, 1981. 191 T. Inoue and M. Nakano, "High accuracy control of a proton synchrotron magnetpowtlrsupply,"IFAC, vol. W p p . 216-221, 1981. [IO] S . Hara, Y. Yamammoto, T. Omata and M. Nakano, "Repetitive control system: a new type servo system for periodic exogenous signals,'' IEEE Trans. Autoinatic Contr., vol. 33. no. 7, pp. 659-66, 1988. [ 111 B. A. Francis and W. M. Wonham, "The internal model principle for linear multivariablr: regulators," Appl. Math. Opt., vol. 2, pp. 170-194, 1975. [12] M. Nakano and S. Hara. "Microprocessor-based repetitive control," Microprocessor-Rased Controi Systems, D. Reidel Piihlishing ('ompany. pp. 279-296. 1986. 1 131 T. Inoue, "Practical repetitive control system design." IEEE Proceedings of the 29th Conference ofDecision and Control, pp. 1673-1678, 1990. [ 141 K. h a n u m a , M. Kozaki, and E ' .Sakasi, "State feedback compensation for repetitive seivo system of dc-dc converter," IEEE f r w . on int Tile. Energy Conf,pp. 268-274, 199 I . 151 T. Haneyoshi, A. Kawamura and R. G. Hoft, "WaveSorm compensation of PWM Inverter with cyclic fluctuating loads," EYE T r a n m c ~ o mon Indu,stryApplication.r, vol. 24, pp.582-589, 1988. 161 Y. Nishida and T Haneyoshi, "Predictive instantaneous value controlled PWM inverter Sor UPS," 23rd Annual IEEE Power Ekctronics Speciahts C'onfgrence, pp.776-783, 1992. 17) K. K. Chew and M. Tomizuka "Steady-state and stochastic performance of a modified discrete-time prototype repetitive controller," IEEE Journal 01Dynamic System.r,Measurement, and Control, vol. 112, pp. 35-41, 1990. [ 181 J. Hu and M. Tomizuka, "A new plug-in adaptive controller for rejection of periodic disturbances," ASME Journal qfDynamrc Systenis.Measurement and Control, vol. 33, pp. 1-5, 199 1. [19] J. Hu and M. Tomizuka, "Adaptive asymptotic tracking of repetitive signals - a frequency domain approach," Proceeding oj'./lmerican Control Conference, vol. 3, pp. 2621-2627,1991. [20] 1,. Ljung, Sy.vtern Identrfkation: Theory for the (iser, Prenirce-Hall. Inc., 1988. 1211 L. Ljung atid T. Soderdtrom, Theory and Praclice o f Recursive Ident$cation, Asco Trade Typeserting Ltd., Hong Kong 1983. 1221 S . Jung, L. H Ho. H. C. Ych. and Y. Y. Tzou, "DSP-hazed digital control of a PWM inverter for sine wave tracking hy optimal state feedback technique," Il'XE I'ESC Con$ Rec.. pp. 546-551, 1994.

plant model of the PWM inverter system. This reveals the feasibility of the construction of an adaptive repetitive control module (ARCM) which can be inserted within the control loop of a conventional analog controlled PWM inverter and significantly improves the waveform quality for ac voltage regulation

REFERENCES
; . Hoft, "rnstantansous feedback controlled PWM [ 11 A. Kawamura and K. < inverter with adaptive hysteresis," IEEE Trans. on Ind. Appl., vol. 112-20, pp. 769-775, J d y / A ~ g 1984. . 121 H. J. Cha, et al. "Real-time digital control of PWM inverter with PI compensator for uninterruptihle power supply," IEEE Ind. Electronics Conj?. vol. 2, pp. 1125-1 128, 199(J. 131 A Kawamura and T. Yokovama. "Comparison of five dificrent approaches for real time digital feedback control of PWM inverter." IEEE Ind. Appl. Sociefit(-on[, vol. 2, pp. 1005- 1000. 1990

163
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY TIRUCHIRAPALLI. Downloaded on May 23,2010 at 08:43:43 UTC from IEEE Xplore. Restrictions apply.