Electric Power Systems Research 96 (2013) 1–8

Contents lists available at SciVerse ScienceDirect

Electric Power Systems Research

journal homepage: [Link]/locate/epsr

Analysis and design of reduced order linear quadratic regulator control for three
phase power factor correction using Cuk rectifiers
M.G. Umamaheswari a,b,∗ , G. Uma a
a
Department of Electrical and Electronics Engineering, College of Engineering, Anna University, Chennai 600 025, Tamil Nadu, India
b
Department of Electronics and Instrumentation Engineering, R.M.K. Engineering College, Kavaraipettai 601 206, Tamil Nadu, India

a r t i c l e i n f o a b s t r a c t

Article history: In this paper, the analysis and design of reduced order linear quadratic regulator (ROLQR) control for
Received 17 February 2012 power factor correction (PFC) in a three phase system is presented. The front end is a three phase diode
Received in revised form rectifier followed by DC–DC Cuk converter modules with the common DC output. Instantaneous symmet-
30 September 2012
rical component theory is used for the generation of reference current. The control strategy uses three
Accepted 17 October 2012
inner ROLQR current controllers for source current shaping and an outer voltage loop using PI controller
Available online 5 December 2012
for load voltage regulation. It uses a single stage converter for both PFC and voltage regulation. The pro-
posed method offers simple control strategy, fast transient response and power factor closed to unity.
Keywords:
Power quality
This type of three-phase three switch PFC converter features a simple and robust configuration compared
Reduced order linear quadratic regulator to conventional six switch topology. To validate the proposed method, a prototype controlled by dSPACE
control 1104 signal processor is set up. Simulation and experimental results indicate that the proposed system
Power factor correction offers regulated output voltage for wide load variations and power factor close to unity.
Proportional integral controller © 2012 Elsevier B.V. All rights reserved.
Total harmonic distortion
Unity power factor

1. Introduction switching losses [6–13]. A three-phase three-switch topology com-

posed of three single-phase single-switch modules was proposed
Recently, there is growing awareness about line pollution and for PFC in [14,15]. Even though the above method offers simple
deteriorating power factor due to the usage of pervading inductive control implementation, it fails to operate in case of one or two
and non-linear loads. Although many solutions were offered for module failures. A derived version of buck boost rectifier is a Cuk
single phase power factor correction (PFC), three phase active PFC rectifier that inverts the voltage polarity and can also simulta-
was seldom considered. As all high power equipment derive elec- neously increase or decrease the voltage magnitude. It has excellent
trical power from three phase mains, incorporating an active three features such as capacitive energy transfer, magnetic components
phase PFC front end can contribute significantly in improving the integrability, full transformer utilization and good steady-state per-
overall power factor and reducing line pollution. Many literature formance. It also provides smooth input and output currents due
have been proposed for PFC. A three-phase single switch PFC topol- to the presence of inductors in the input and output side [16].
ogy has the merits of simple control and few components [1–5]. Many methods for generating the reference template were pro-
This type of converter suffers due to Discontinuous Conduction posed. In [17–19] the instantaneous reactive power theory was
Mode (DCM) operation, causing high current stresses on the power proposed (i.e. p–q theory) for calculating the reference currents.
devices. A three phase six switch Pulse Width Modulation (PWM) The general equation for deriving the reference current which
boost rectifier is widely used for high-power applications. Advan- relates the new concept of instantaneous active and reactive theory
tages of this type of rectifier are high efficiency and good current was reported in [20], but no detailed information was given for DC
quality. But the PWM rectifier needs a complex control structure bus voltage compensation. In [21], the extended p–q theory was
and the efficiency is lower than the diode rectifier due to additional proposed to derive a more general vector equation for calculating
the reference currents. This method does not operate satisfactorily
for module loss. In [22], only the final formulation of the extracted
reference current was reported. In [23] and [24], PFC using Cuk
∗ Corresponding author at: Department of Electrical and Electronics Engineering,
rectifier modules was proposed and reference current was gen-
College of Engineering, Anna University, Chennai 600 025, Tamil Nadu, India.
erated using power balance control technique. The extraction of
Tel.: +91 8148747320.
E-mail addresses: umagopinathh@[Link] (M.G. Umamaheswari), reference current based on instantaneous symmetrical component
uma@[Link] (G. Uma). theory involves simple computations based on the instantaneous

0378-7796/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
[Link]
2 M.G. Umamaheswari, G. Uma / Electric Power Systems Research 96 (2013) 1–8

source voltages and currents and it does not require any three phase 2.1. Design of DC–DC Cuk converter components
to two phase conversions.
The classic Linear Quadratic Regulator (LQR) approach deals A Cuk PFC was designed [23] with the following specifications:
with the optimization of a cost function or performance index. Thus, rectified input voltage Vg = 24 V, line frequency 50 Hz, power fac-
the designer can weigh which states are more important in the con- tor ≥0.99, maximum output power Po = 56 W, RL = 10 , output DC
trol action to seek for appropriate performance. This feature of LQR voltage Vo = − 24 V, efficiency  ≥ 85%. DC voltage conversion ratio
control has initiated several researchers to successfully apply this M is given by
technique in the field of power electronics. In [25] and [26], the
Vo
performance indices are selected using pole placement technique. M= = 0.704 (6)
Vg
This method depends on the exact placement of closed loop poles.
In [27], the cost function is derived from an initial controller by Conduction parameter of the Cuk PFC circuit
frequency domain method which is a time consuming procedure. 1
Optimal control provides a systematic way of designing a LQR con- Ka >   2 > 0.1752 (7)
troller and problem of where to place the closed loop poles does (2(M + sin()) )
not arise and is robust to parameter and load variations.
An equivalent inductance is given by
Hence, it is proposed to develop a single stage three phase
AC–DC converter using Cuk rectifier modules based on ROLQR con- RL TS Ka
Leq = = 0.667 mH (8)
trol for achieving voltage regulation and PFC. The instantaneous 2
symmetrical component theory is used for calculating the refer- By choosing the input inductor current with 6.75% current rip-
ence currents. To reduce the complexity in analysis, a reduced order ple, iLi = 0.169 A. The design of Li and Lo is made using the desired
model of Cuk converter is obtained from the higher order state ripple value of the input current. For the duty ratio d = 0.5, switching
space model [28,29] by Pade’s approximation technique. The outer frequency = 50 kHz, switching period Ts = 0.25 ␮s, the input induc-
voltage loop compensator is designed using Ziegler Nichol’s tuning tor value is found to be
procedure.
Vg dTS
Li = = 2 mH (9)
iLi
2. Modeling of DC–DC Cuk converter
The output inductor value is found to be

Fig. 1(a) represents the circuit diagram of the Cuk converter. It L1 Leq
Lo = = 1 mH (10)
consists of two inductors Li , Lo and two capacitors Ct , Co , Vg and L1 − Leq
Vo represent supply and output voltage, respectively, S is an active
Transfer capacitor value is calculated by considering a resonant
switch, D is a freewheeling diode and RL is the load resistance. S
frequency of 1 kHz and is given by
operates at a switching frequency fs with duty ratio d.
To obtain the mathematical model of the controller, the state 1
Ct = 2 (L − L )
= 25.35 ␮F (11)
model of Cuk converter is derived by considering S = 1 during the ωro i o
MOSFET switch conduction subinterval and S = 0 during the diode
To select the output capacitance value in this circuit, the main
conduction subinterval. The converter dynamics is described by
factor is the output ripple voltage which is caused by the second
state–space averaging method and by using the same method, the
harmonic. The equation to define the output capacitance value can
state equations during switch-on and switch-off conditions are
be expressed as
combined as follows:
Pin
dx1 (1 − d) Vg Co = = 2211 ␮F (12)
=− x3 + (1) 2fr Vo Vo
dt Li Li
where Vo is the peak value of the output ripple voltage = 0.6 V
dx2 d 1 (2.5% of the output voltage) and fr is the second harmonic frequency.
= x3 − x4 (2)
dt Lo Lo The design values are substituted in (5) and after applying phase
variable transformation, A and B matrices become,
dx3 (1 − d) d ⎡ ⎤ ⎡ ⎤
= x1 − x2 (3) 0 1 0 0 0
dt Ct Ct
⎢
A=⎣
0 0 1 0 ⎥, ⎢0⎥
dx4 1 1 0 0 0 1
⎦ B=⎣ ⎦
0
(13)
= x2 − x4 (4)
dt Co RL C0 −6.3 × 1012 −1.87 × 109 −3.8 × 107 −50 1

where x1 , x2 , x3 , x4 are the current through the inductor (iLi ), cur- The main objective is to shape the input source current and to regu-
rent through the inductor (iLo ), voltage across the transfer capacitor late the output voltage of the DC–DC Cuk converter. The converter
(vct ), voltage across the output capacitor (vco ), respectively, and d requires sensing of four state variables, which is not acceptable
represents the duty cycle. From Eqs. (1)–(4), the averaged system from practical point of view. In order to reduce the complexity
matrices are derived as given below: in controller design, the fourth order model of a Cuk converter is
reduced to a second order model. The model reduction technique
⎡ ⎤
−(1 − d) used is Pade’s approximation, wherein the two dominant poles of
0 0 0 ⎡ ⎤
⎢ Li ⎥ the system i.e. the input inductor current, iLi and the output capaci-
⎢ ⎥ 1
tor voltage, vCo are retained and the effects of the transfer capacitor
⎢ d 1 ⎥ ⎢ L ⎥
⎢ 0 0 − ⎥ ⎢ i ⎥ Ct , and the output inductor Lo are neglected. Hence it becomes suffi-
⎢ Lo L0 ⎥ ⎢0 ⎥
A=⎢ ⎥ , B = ⎢ ⎥ (5) cient to regulate these two variables iLi , vCo using the ROLQR control
⎢ (1 − d) d ⎥ ⎣0 ⎦
⎢ − 0 0 ⎥ strategy. The reduced order matrix for the above system is given as
⎢ Ct Ct ⎥
⎣ ⎦ 0 0 1 0
1 1 A= , B= (14)
0 0 − −1.64 × 105 −49.34 1
Co RL Co
 M.G. Umamaheswari, G. Uma / Electric Power Systems Research 96 (2013) 1–8 3

Fig. 1. (a) Circuit diagram of DC–DC Cuk converter. (b) Circuit diagram of single phase Cuk PFC rectifier. (c) Circuit diagram of the three phase Cuk PFC rectifier module.

2.2. Cuk rectifier module vsb + (vsc − vsa )ˇ∗

isb = (PLoad + PLoss ) (16)
v2sa + v2sb + v2sc
The circuit diagram of the single phase Cuk PFC is shown in
Fig. 1(b). The converter is designed using the design values out- vsc + (vsa − vsb )ˇ∗
isc = (PLoad + PLoss ) (17)
lined in the previous section. The circuit diagram of the proposed v2sa + v2sb + v2sc
three phase configuration using single phase topology is shown in
√
Fig. 1(c). Lia , Lib , Lic represent the input inductors and Loa , Lob , Loc where e=Vref − Vo and ˇ∗ = tan ϕ/ 3, ϕ is the desired phase angle
represent the output inductors for the respective three phases a, between supply voltages (vsa , vsb , vsc ) and source line currents (isa ,
b, c respectively. Cta , Ctb , Ctc represent the transfer capacitors. vsa , isb , isc ) for the balanced system. Here ϕ is taken as zero and PLocal is
vsb , vsc and vo represent supply voltages for the three phases and the load power and is given by
output voltage, respectively, Sa , Sb , Sc are active switches for the
three phases, Da , Db , Dc are freewheeling diodes and RL is the load PLoad = Vo × ILoad (18)
resistance. All power devices of the PFC circuit, including the diode
rectifier and the power circuit components of the Cuk converter Here the term PLoss represents the switching losses and ohmic
are modularized. A single capacitor Co is connected at the output losses in the converter. Due to losses in converter, the capacitor
terminals for filtering the output voltage ripple. voltage will decrease. When the capacitor voltage falls below the
reference voltage; the PFC rectifier may not be able to track the
reference current faithfully. So a suitable PI controller is used which
2.3. Generation of reference currents
regulate the voltage of capacitor to the reference value. PLoss is thus
given by
The theory of instantaneous symmetrical components can com-
pensate any kind of harmonics under balanced supply conditions.
Reference current generation based on symmetrical component PLoss = Kp e + Ki edt (19)
theory is explained in Fig. 2. Based on this theory, the reference
current for the three phase balanced supply is given by
3. Control scheme
vsa + (vsb − vsc )ˇ∗
isa = (PLoad + PLoss ) (15)
2 + v2 + v2
vsa This section presents the design of the controller. The controller
sb sc
is comprised of three inner ROLQR current controllers for shap-
ing the source current and an outer voltage control loop using PI
controller to regulate the output voltage as shown in Fig. 3.

3.1. Principle of PI controller

A Proportional Integral (PI) compensator is selected for pro-

viding DC voltage regulation. The output voltage Vo is sensed and
compared with reference voltage Vref and the error is processed by
the PI controller to keep the output voltage constant. The values of
Kp and Ki are found by Ziegler–Nichols tuning method. The tuning
method provides sustained oscillations with the ultimate gain
Ku = 4 and ultimate period Pu = 0.065 s. By this technique, the exact
values for Kp (=Ku /2) and Ki are found to be 2 and 18.46 (Ti = Pu /1.2,
Fig. 2. Implementation of Instantaneous Symmetrical Component Theory. Ki = 1/Ti ), respectively. The closed loop operation of proposed
4 M.G. Umamaheswari, G. Uma / Electric Power Systems Research 96 (2013) 1–8

Fig. 3. Implementation of ROLQR control.

Fig. 4. Simulated waveforms for ROLQR control (a). Three phase source currents (b). UPF operation in phase-a (c). Regulated output voltage (d). Source currents with module
loss (e). Regulated voltage and load current for step change in load (f). Spectrum of source current in phase-a.
 M.G. Umamaheswari, G. Uma / Electric Power Systems Research 96 (2013) 1–8 5

For the proposed system, the quadratic cost function is obtained

after substituting the values of X, Q, R in Eq. (21)
tf

[(1.16 × 107 x12 + x22 ) + u2 ]dt (23)

t0

The control law is found to be

u = −R−1 BT Kx = u = −KF x (24)

where KF is the feedback gain matrix and K is the return function

matrix. The unknown coefficients of the return function matrix are
found by solving the Ricatti equation

−Q − AT K − KA + KBR−1 BT K = 0 (25)
    
7 5
1.16 × 10 0 0 −1.64 × 10 k11 k12
− −
0 −49.34 k21 k22
  1 1
k11 k12 0 1
− + (26)
k21 k22 −1.64 × 105 −49.3
   
k11 k12 0   k11 k12
[1] 1 0 =0
k21 k22 1 k21 k22

On solving Eq. (26), the return function matrix K is found to be

   
k11 k1 1.2 × 105 40
= (27)
Fig. 5. Simulation results for the source currents, inductor currents, output voltage k21 k22 40 1
and load current and output power for symmetric mains condition, module loss in
phase a, module loss in phases a and b for ROLQR control. On substituting the value of K matrix in the equation,
KF = −R–1 BT K the feedback gain matrix KF is obtained. It is found to
be 35.362 0.7216 .
system using outer PI controller and inner ROLQR controller is
Hence the control law becomes
depicted in Fig. 3.
u = −35.362 (x1ref − x1 ) − 0.7216 (x2ref − x2 ) (28)
3.2. Reduced order linear quadratic regulator control Eq. (28) is rewritten as

The theory of optimal control is concerned with operating u = −(K1 e1 + K2 e2 ) where K1 = 35.362 and K2 = 0.7216 (29)
a dynamic system at minimum cost. The time invarying linear
quadratic regulator is used as tracking current controller. Here for 3.2.1. Closed loop control using ROLQR
the proposed system ROLQR gain matrix is computed by appropri- The proposed ROLQR scheme is illustrated in Fig. 3, where PI
ately choosing values of R and Q (weight matrix). The values for Q controller and ROLQR act as outer voltage controller and inner cur-
and R matrices are rent controllers, respectively. The output voltage is compared with
  the reference voltage and the error e1 is fed to the PI controller.
1.16 × 107 0 Reference currents are generated using the output of the PI con-
Q = , R = [1] (20)
0 1 troller, i.e. PLoss , three phase source voltages vsa , vsb , vsc and the
load power PLoad. The reference currents with desired magnitude
Q matrix is chosen in such a way that maximum weightage is and shape are derived by instantaneous symmetrical component
given to input inductor current, so that input current shaping is theory. The instantaneous values of actual current and reference
performed more effectively by ROLQR controller. Q and R should current are compared and the error is found to be e2. The inputs
be positive semi-definite and positive definite, respectively. They to the ROLQR are voltage error e1 and the current error e2 for each
are selected such that the scalar quantity xT Qx is always positive or phase. The output u is the control signal in each phase, which in
zero at each time t for all functions x(t), and the scalar quantity uT turn sets the new duty ratio of the switching pulses for triggering
Ru is always positive at each time t for all values of u(t). In terms of the switches Sa , Sb , Sc .
eigen values, the eigen values of Q should be non-negative, while
those of R should be positive. 4. Simulation studies
For a continuous time system, the state-feedback control law
u = −KF x minimizes the quadratic cost function A simulation model of the three phase AC–DC converter formed
by single phase Cuk rectifier modules controlled using PI controller
tf
for output voltage regulation and ROLQR for UPF operation is devel-
1
J(x(·), u(·)) = (xT Qx + uT Ru)dt (21) oped in MATLAB/SIMULINK. The entire system is simulated. The
2
simulation results of the proposed scheme using ROLQR are shown
t0
in Fig. 4(a)–(e). It can be seen from Fig. 4(a) and (b) that almost unity
Subject to the system dynamics power factor and nearly sinusoidal current operation is achieved. It
is understood from Fig. 4(c) that the proposed system also provides
X • = AX + BU (22) regulated output voltage. Fig. 4(d) shows the source currents under
6 M.G. Umamaheswari, G. Uma / Electric Power Systems Research 96 (2013) 1–8

Fig. 6. Experimental waveforms for ROLQR control (a). Three phase source currents (b). UPF operation in phase-a (c). Regulated output voltage (d). Source currents with
module loss (e). Regulated voltage for step change in load (f). Spectrum of source current in phase-a.

module loss conditions. All the three-phase supply condition exists inductor currents, output voltage, load current and output power
between 0.29 and 0.34 s and module loss occurs in phase b between for symmetric mains condition, module loss in phase a, module loss
0.34 and 0.41 s: and module loss occurs in phases b and c between in phases a and b.
0.41 and 0.51 s. The simulation result of transient response of the
output voltage and load current waveforms for step load change 5. Experimental studies
is shown in Fig. 4(e) and it is observed that the output voltage
is maintained constant for a step load change. Here, each single A prototype is constructed in order to experimentally test
phase PFC Cuk module is designed for the rated load current. In the the ROLQR control method using the following specifications
three phase configuration, each Cuk module carries one-third of fs = 50 kHz, Vs = 24 Vrms , Lia , Lib , Lic = 2 mH, Loa , Lob , Loc = 1 mH, Cta , Ctb ,
load current. Hence, from Fig. 4(d), it is observed that, even under Ctc = 25 ␮F, Co = 2000 ␮F, RL = 10  using dSPACE 1104 signal pro-
single/two module loss conditions, the system continues to work cessor. The input and output inductors are of ferrite core type and
and maintain the output voltage nearer to the desired value. Under the capacitors are of plain polyester type. Power MOSFET IRF540N
one module loss conditions, the magnitude of the source currents is used as a switch and IN 4007 is used as a diode. The 12 bit ADC unit
are unequal and it is due to the presence of an oscillating compo- in the dSPACE accepts a maximum analog input voltage of ±10 V.
nent in the power PLavg. From Fig. 4(f), the %THD is found to be During the closed loop operation the three phase source voltages
3.26. Fig. 5 shows the simulation results of the input line currents, and source currents, output voltage and current are sensed by hall

Table 1
A summary of system performance parameters for the variation of load.

Load current (A) 2.404 2.186 2.004 1.85 1.718 1.603

Efficiency (%) 86.79 86.49 86.21 85.73 85.34 84.88
PF 0.9921 0.9919 0.9919 0.9919 0.9918 0.9916
THD 2.311 2.847 3.203 3.521 4.102 4.532
Voltage regulation (%) −0.166 −0.249 −0.249 −0.249 −0.249 −0.166
 M.G. Umamaheswari, G. Uma / Electric Power Systems Research 96 (2013) 1–8 7

Fig. 7. Graphical representation of data given in Table 1 illustrates the system performance as the load is varied.

effect voltage and current sensors and scaled down to less than 7. Conclusion
±10 V using the signal conditioning circuit.
The ADC signals are processed by the proposed ROLQR algorithm In this paper, the analysis and design of a single stage three
to calculate the new duty ratio for the switches Sa , Sb , Sc . The PWM phase AC to DC converter formed by DC Cuk rectifier modules with
pulses obtained from dSPACE unit is used to trigger the MOSFET the common DC output for PFC was presented. Reference current
switches Sa , Sb , Sc using IR2110 driver. Fig. 6 depicts the experimen- was generated using instantaneous symmetrical component the-
tal results of the proposed converter module. Fig. 6(a) reveals that ory .The control strategy is based on outer voltage control loop and
source current becomes sinusoidal after implementing ROLQR con- three inner ROLQR current controllers. To support the proposed
trol. UPF operation at the line side under constant load conditions, method, a prototype controlled by dSPACE signal processor was set
tight regulation of the DC output voltage is evident from Fig. 6(b) up. The variation of performance parameters of Cuk rectifier mod-
and (c). Fig. 6(d) shows the source currents under module loss con- ules with the load variation proves the robustness and effectiveness
ditions. Fig. 6(e) shows that the converter gets back its reference of ROLQR control. Simulation and experimental results reveal that
voltage for step change in load. The closed loop control of Cuk rec- the proposed system offers the regulated output voltage for step
tifier module using ROLQR forces the source current to follow the load variations and also provide power factor close to unity.
source voltage more effectively and it is also proved by viewing the
harmonic spectrum of source current which is shown in Fig. 6(f)
and the %THD is found to be 4.3%. Table 1 presents the variation References
of performance parameters as a function of load current and the
[1] H.-J. Chiu, T.-H. Wang, L.-W. Lin, Y.-K. Lo, Current imbalance elimination for a
performance curves are depicted in Fig. [Link] Fig. 7, it is observed
three-phase three-switch PFC converter, IEEE Transactions on Power Electron-
that the % efficiency is maintained around 85% for all the loads and ics 23 (March (2)) (2008) 1020–1022.
the maximum efficiency at rated load is found to be 86.79% at nomi- [2] R.L. Alves, I. Barbi, Analysis and implementation of a hybrid high-power-factor
nal load current of 2.404 A. The %THD is almost maintained constant three-phase unidirectional rectifier, IEEE Transactions on Power Electronics 24
(March (3)) (2009) 632–640.
and the power factor is also maintained around unity irrespective of [3] G. Spiazzi, F.C. Lee, Implementation of single-phase boost power-factor-
the load variation which shows the effectiveness of ROLQR control. correction circuits in three-phase applications, IEEE Transactions on Industrial
Electronics 44 (June (3)) (1997) 365–371.
[4] A.R. Prasad, P.D. Ziogas, S. Manias, An active power factor correction tech-
nique for three-phase diode rectifiers, IEEE Transactions on Power Electronics
6. Discussion 6 (January (1)) (1991) 83–92.
[5] L. Dalessandro, S.D. Round, J.W. Kolar, Center-point voltage balancing of hys-
teresis current controlled three-level PWM rectifiers, IEEE Transactions on
It is proposed to construct a ROLQR based single stage Cuk Power Electronics 23 (September (5)) (2008) 2477–2488.
converter in modular form for achieving PFC and voltage regula- [6] C. Qiao, K.M. Smedley, Unified constant-frequency integration control of three-
tion. All the control circuits including the PI controller, reference phase standard bridge boost rectifiers with power-factor correction, IEEE
Transactions on Industrial Electronics 50 (February (1)) (2003) 100–107.
current generation using instantaneous symmetrical component [7] S. Chattopadhyay, V. Ramanarayanan, Digital implementation of a line current
theory and inner ROLQR control loop have been implemented using shaping algorithm for three phase high power factor boost rectifier with out
dSPACE signal processor. There is a difference in %THD between input voltage sensing, IEEE Transactions on Power Electronics 19 (May (3))
(2004) 709–721.
simulation and experimentation, because the internal resistances [8] S. Chattopadhyay, V. Ramanarayanan, A voltage-sensorless control method to
of reactive elements, semiconductors that may slightly influence balance the input currents of a three-wire boost rectifier under unbalanced
the dynamic of the converter, which is not appearing when simu- input voltages condition, IEEE Transactions on Industrial Electronics 52 (2)
(2005) 386–398.
lated with ideal components. Moreover the power quality analyzer
[9] R. Ghosh, G. Narayanan, Control of three-phase, four-wire PWM rectifier, IEEE
used in experimentation is able to trap only up to 49th harmonic. Transactions on Power Electronics 23 (January (1)) (2008) 96–106.
8 M.G. Umamaheswari, G. Uma / Electric Power Systems Research 96 (2013) 1–8

[10] X.H. Wu, S.K. Panda, J.X. Xu, DC link voltage and supply-side current harmonics [19] S.J. Huang, J. Chang-Wu, A new control algorithm for three phase three wire
minimization of three phase PWM boost rectifiers using frequency domain active filters under non ideal main voltages, IEEE Transactions on Power Elec-
based repetitive current controllers, IEEE Transactions on Power Electronics tronics 14 (July (4)) (1999) 3089–3094.
23 (July (4)) (2008) 1987–1997. [20] M. Aredes, E.H. Watanabe, New control algorithms for series and shunt three
[11] P. Xiao, K.A. Corzine, G.K. Venayagamoorthy, Multiple reference frame-based phase four wire active filter, IEEE Transactions on Power Delivery 10 (July (3))
control of three-phase PWM boost rectifiers under unbalanced and distorted (1995) 1649–1656.
input conditions, IEEE Transactions on Power Electronics 23 (July (4)) (2008) [21] M.K. Mishra, A. Joshi, A. Ghosh, A new compensation algorithm for balanced
2006–2017. and unbalanced distribution systems using generalized instantaneous reactive
[12] A.H. Bhat, P. Agarwal, Three-phase, power quality improvement ac/dc convert- power theory, Electric Power Systems Research 60 (2001) 29–37.
ers, Electric Power Systems Research (2007) 276–289. [22] C.C. Chen, Y.Y. Hsu, A novel approach to the design of a shunt active filter for
[13] Y. Neba, K. Ishizaka, R. Itoh, Single-phase two-stage boost rectifiers with sinu- unbalanced three phase four wire systems under non sinusoidal conditions,
soidal input current, IET Power Electronics 3 (March (2)) (2010) 176–186. IEEE Transactions on Power Delivery 15 (October (4)) (2000) 1258–1264.
[14] Y.K. Eric Ho, S.Y.R. Hui, Y.-S. Lee, Characterization of single-stage three-phase [23] U. Kamnarn, V. Chunkag, Analysis and design of a modular three phase AC-to-
power-factor-correction circuit using modular single-phase PWM DC-to-DC DC converter using CUK rectifier module with nearly unity power factor and
converters, IEEE Transactions on Power Electronics 15 (January (1)) (2000) fast dynamic response, IEEE Transactions on Power Electronics 24 (August (8))
62–71. (2009) 2000–2012.
[15] S. Kim, P.N. Enjeti, A modular single-phase power-factor-correction scheme [24] V. Chunkag, U. Kamnarn, Paralleling three-phase AC to DC converter using
with a harmonic filtering function, IEEE Transactions on Industrial Electronics CUK rectifier modules based on power balance control technique, IET Power
50 (April (2)) (2003) 328–335. Electronics 3 (July (4)) (2010) 511–524.
[16] M.G. Umamaheswari, G. Uma, K.M. Vijayalakshmi, Design and implemen- [25] F.H.F. Leung Peter, K.S. Tam, C.K. Li, An improved LQR-based controller for
tation of reduced-order sliding mode controller for higher order power switching DC–DC converters, IEEE Transactions on Industrial Electronics 40
factor correction converters, IET Power Electronics 4 (November (9)) (2011) (October (5)) (1993) 521–528.
984–992. [26] F.H.F. Leung, P.K.S. Tam, C.K. Li, The control of switching DC–DC converters – A
[17] H. Akagi, Y. Kanazawa, A. Nabae, Instantaneous reactive power compen- general LQR problem, IEEE Transactions on Industrial Electronics 38 (February
sators comprising switching device without energy storage components, (1)) (1991) 65–71.
IEEE Transactions on Industry Applications IA-20 (May/June (3)) (1984) [27] C. Gezgin, B.S. Heck, R.M. Bass, Control structure optimization of a boost con-
625–630. verter: an LQR approach, in: Proceedings of IEEE PESC, vol. 2, 1997, pp. 901–907.
[18] F.Z. Peng, J.-S. Lai, Generalized instantaneous reactive power theory for three- [28] S. Ang, et al., Power Switching Converter, Taylor and Francis Group, 2005.
phase power systems, IEEE Transactions on Instrumentation and Measurement [29] J.S. Bay, Fundamentals of Linear State Space Systems, McGraw-Hill Publishers,
45 (February (1)) (1996) 293–297. 1998.