performance analysis of d-statcom ... analysis of d-statcom compensator using control techniques for...

PERFORMANCE ANALYSIS OF D-STATCOM

COMPENSATOR USING CONTROL TECHNIQUES

FOR LOAD COMPENSATION

Mr.Ch. VENKATA KRISHNA REDDY, Dr. K.KRISHNA VENI,

Dr.G.TULASIRAM DAS and Dr. SAMPATH.PULLA

ABSTRACT

The majority of power consumption has been drawn in reactive loads, these

excessive reactive power demand increases feeder losses and reduces the active

power flow capability of distribution system where as unbalancing affects the

operation of transformers and generators. DSTATCOM can be used for the

compensation of reactive power and unbalance loading in distribution system.

The performance of the DSTATCOM depends on the control algorithm i.e. the

extraction of the current components. In this project DSTATCOM is controlled

by IRP and SRF theory for compensation of reactive power and unbalance and

these methods are compared with a new Adaline based algorithm. Adaline based

control algorithm An Adaline based control technique has resulted

inconsiderable improved performance of DSTATCOM.

INDEX TERMS — DSTATCOM, Extraction of current components,

Reactive power, Control algorithms, Matlab.

I. INTRODUCTION

A power system is said to be well designed only if it provides good voltage

profile. In the early days of power transmission in the late 19th century problems

like voltage deviation during load changes and power transfer limitation were

International Journal of Electrical and Electronics Engineering Research (IJEEER) Vol.1, Issue 2 Dec 2011 149-171 © TJPRC Pvt. Ltd.,

Mr. Ch. Venkata Krishna Reddy, Dr. K. Krishna Veni , Dr. G. Tulasiram Das and Dr. Sampath.Pulla

150

observed due to reactive power unbalances. Today these Problems have even

higher impact on reliable and secure power supply in the World of Globalization

and Privatization of electrical system and energy transfer. The development in

fast and reliable semiconductors devices (GTO and IGBT) allowed new power

electronic Configurations to be introduced to the tasks of power Transmission

and load flow control. The FACTS devices offer a fast and reliable control over

the transmission parameters, i.e. Voltage, line impedance, and phase angle

between the sending end voltage and receiving end voltage. On the other hand

the custom power is for low voltage distribution, and improving the poor

quality and reliability of supply affecting sensitive loads. Custom power devices

are very similar to the FACTS. Most widely known custom power devices are

DSTATCOM, UPQC, DVR among them DSTATCOM is very well known and

can provide cost effective solution for the compensation of reactive power and

unbalance loading in distribution system.

The performance of the DSTATCOM depends on the control algorithm i.e.

the extraction of the current components. In this paper we will discuss about

instantaneous reactive power (IRP) theory, synchronous reference frame (SRF)

theory and compared with the new algorithm ADALINE based on neural

network.

II. SYSTEM CONFIGURATION

Fig. 1 shows the basic circuit diagram of the DSTATCOM system with

lagging power factor loads connected to 3-phase, 3-wire distribution system. The

lagging power-factor load is realized by star connected resistive inductive (R-L)

load. Unbalanced load is realized by disconnecting load from phase-a using

circuit breaker. The three-phase voltage source converter (VSC) working as a

DSTATCOM is realized using six IGBT switches with anti-parallel diodes. At ac

side the interfacing inductors are used to filter out the high frequency

components of the compensating currents.

Performance Analysis of D-STATCOM Compensator using Control Techniques for Load Compensation

151

Fig. 1 : DSTATCOM connected to 3-Phase System

The DSTATCOM has been developed by Westinghouse as a part of EPRI

(Electric power research institute) custom power program for advanced

distribution system. DSTATCOM is a solid state dc-ac switching power

converter that consists of 3-phase voltage source forced air-cooled inverter based

static compensator connected in shunt with the system. In its basic form

DSTATCOM injects a voltage in phase with system voltage, thus providing

voltage support and regulation of VAR flow .The DSTATCOM is capable of

generating continuously variable inductive or capacitive shunt compensation at a

level up its maximum MVA rating. The DSTACOM is available in rating from 2

to 10MVA in modular 2-MVA increments. DSTATCOM can also behave as a

shunt active filter, to eliminate unbalance or distortions in the source current or

the supply voltage as per the IEEE-519 standard limits.

III. CONTROL ALGORITHMS

The performance of the DSTATCOM depends on the control algorithm i.e.

the extraction of the current components. For this purpose there are many control

schemes which are reported in the literature and some of these are instantaneous

compensation, instantaneous symmetrical components, synchronous reference


152

frame (SRF) theory, computation based on per phase basis, and scheme based on

neural network. Among these control schemes instantaneous reactive power

theory and synchronous rotating reference frame are most widely used.

For reactive power compensation, DSTATCOM provides reactive power as

needed by the load and therefore the source current remains at unity power factor

(UPF). Since only real power is being supplied by the source, load balancing is

achieved by making the source reference current balanced. The reference source

current used to decide the switching of the DSTATCOM has real fundamental

frequency component of the load current which is being extracted by these

techniques.

A. Instantaneous Reactive Power Theory

Instantaneous reactive power theory has been initially proposed by Akagi in

1983 The theory is based on the transformation of three phase quantities to two

phase quantities in α –β frame and calculation of instantaneous active and

reactive power in this frame.

INSTANTANEOUS P-Q THEORY

The p-q theory uses 0−− βα transformations and various definitions of

active and reactive powers. The transformation of 3-Phase quantities (a-b-c) of

stationary reference frame into 2-Phase quantities ( 0−− βα ) stationary

reference frame is achieved by applying Clark’s transformation.

1 1 12 2 2

02 1 113 2 2

3 302 2

v vav vbv vc

αβ

= − −

−

(1)


153

1 1 12 2 2

02 1 113 2 2

3 302 2

i iai ibi ic

αβ

= − −

−

(2)

Let the instantaneous real, imaginary and zero sequence power denoted by p,

q and 0p respectively. Then the three phase instantaneous power is given by

3P v i v i v ia a c cb bφ = + + (3)

Using the transformation ( )3 0 0P t v i v i v iα αφ β β= + +

( ) ( ) ( )3 0p t p t p tφ = + (4)

The zero sequence components of the voltages and currents do not

contribute to the instantaneous powers p and q .The instantaneous imaginary

power, q is defined as (5)

( ) ( ) ( ) 1

3q v i v i i v v i v v i v va c b b a c c b aα αβ β= − = − + − + − (5)

Fig. 2 : Block Diagram of IRP theory


154

It represents an energy that may be constant or not and that is being

exchanged between the phases of the system. This implies that q does not

contribute to the energy transfer between the source and the load at any time.

From Eq. (2.20)-(2.21), we write p and q as following.

v v ip

iv vq

α αββαβ

= −

(6)

The instantaneous active and reactive powers p and q can be decomposed

into an average and an oscillatory component

P=P- +P~

q=q- +q~

P_= Mean value of the instantaneous real power. It corresponds to the energy per

time unity that is transferred from the power source to the load, in a balanced

way, through the a-b-c coordinates (it is, indeed, the only desired power

component to be supplied by the power source).

P~ = Alternating value of the instantaneous real power. It is the energy per time

unity that is exchanged between the power source and the load, through the a-0-c

coordinates. Since p” does not involve any energy transference from the power

source to load, it must be compensated.

Therefore the reference source currents i sα and i sβ in α-β coordinate are

expressed as:

The instantaneous currents on the α β− coordinates, iα and

iβ are divided

into two kinds of instantaneous current components respectively.


155

Finally, through o−− βα inverse transformation we compute the filter

reference currents in cba −− phase system

1/ 2 1 02 1/ 2 1/ 2 3 / 23

1/ 2 1/ 2 3 / 2

i ifa o

i ifb f

i ifc f

α

β

∗−

∗ ∗= −

∗ ∗− − (7)

Where io is the zero sequence component which is zero in 3- phase 3-wire

system. The extracted current components are followed by hysteresis based

PWM current controller to generate switching pluses and these signals are

applied to DSTATCOM

The p-q theory has the following limitations

♦ The source voltages are unbalanced and non-sinusoidal.

♦ It needs a large number of transducers of measurements and intensive

computation. It tends to make the system operation complex.

Disadvantages of P-Q theory

• Delay in computation due LPF used for filtering power signals

• This theory uses volt signals to compute instantaneous active and

reactive powers, if any variations in supply voltage causes inaccurate

measurements

• It is not applicable for single phase systems and three phase systems

with neutral.

• Complete harmonic suppression is not achieved in case of nonlinear

loads.


156

• A larger filter inductance is required to filter the harmonics

B. Synchronous rotating frame theory

The synchronous reference frame theory is based on the transformation of

the currents in synchronously rotating d-q frame. Fig.3. explains the basic

building blocks of the theory. If θ is the transformation angle, then the currents

transformation from α-β to d-q frame is defined as:

Fig. 3 : Block Diagram of SRF theory

The 3-Phase load currents of system i.e. La , Lb, Lc,i i i of stationary

reference system in a-b-c coordinates are transformed to two phase system with

α-β coordinates, this can be done by algebraic transformation, known as “Clark’s

transformation” which also produces a stationary reference system again these

currents are transformed to synchronously rotating reference frame by Park’s

transformation. The voltage templates (sine and cosine) are generated using

phase locked loop (PLL) and then we calculate the d , q ,i i which are passed

through LPF which filter out the harmonics in current signals and again these two

phase currents in rotating reference frame are transformed to two phase

stationary reference frame using reverse park’s transformation and then

transformed to three phase coordinates which are called as the reference current

signals


157

−=

ii

ii

q

d

β

α

θθθθ

cossin

sincos

(8)

SRF isolator extracts the dc component by low pass filters (LPF) for each id

and iq realized by moving averager at 100Hz. The extracted DC components

iddc and iqdc are transformed back into α-β frame as shown below:

−=

ii

ii

qdc

ddc

dc

dc

θθθθ

β

α

cossin

sincos

(9)

From here the transformation can be made to obtain three phase reference

currents in a-b-c coordinates using Reverse Park’s transformation and Clark’s

transformation. The reactive power compensation can also be provided by

keeping iq component zero for calculating reference currents.

Disadvantages of SRF theory:

• The generation of voltage templates (sin and cosine terms is crucial)

• PLL (phase locked loop) is used generation of templates. The operation

of PLL is slow and also imposes some of amount of delay in

computation

• Effect of delay due to LPF used for filtering signals in d- q frame

• Mathematical computations are more and not suitable for 3-phase 4-

wire system.

C. Adaline based control algorithm

ADALINE (Adaptive Linear Element) is a single layer neural network. It

was developed by Professor Bernard Widrow and his graduate student Ted


158

Hoff at Stanford University in 1960. It is based on the McCulloch–Pitts neuron.

It consists of a weight, a bias and a summation function.

The basic theory of the proposed decomposer has been based on Least Mean

Square (LMS) algorithm and its training through Adaline, which tracks the unit

vector templates to maintain minimum error. The block diagram of Adaline

based control algorithm is given in Fig 4.. The basic concept of theory used here

can be understood by considering the analysis in single-phase system which is

given as under

The supply voltage may be expressed as:

Vs=V1Sinwt (10)

Load current is made up of active current (ip), reactive current (iq) for

positive sequence and negative sequence current (i−) can be decomposed in parts

as:

The control algorithm is based on the extraction of current component in

phase with the unit voltage template. To estimate fundamental frequency positive

sequence real component of load current, the unit voltage template should be in

phase with the system voltage and should have unit amplitude further it must be

undistorted.


159

Fig. 4 : Block diagram of the ADALINE theory

For calculation of templates the voltage at Pcc is sensed. The sense voltage

is filtered through a band pass filter and the instantaneous rms value is computed.

The sensed three phase voltages are divided by this instantaneous rms value to

get three phase voltage templates.

The initial estimates of active part of current for single phase can be chosen

as:

i p = W p u p (11)

Where weight (Wp) is estimated using Adaline. The weight is variable and

changes as per the load current and magnitude of phase voltage. Algorithm for

estimating weights corresponding to fundamental frequency real component of

current (for three phase system), based on LMS algorithm tuned Adaline tracks

the unit voltage templates to maintain minimum error L(K) p(K) p(K)i w u−

P ( k + i ) = WP ( k ) +пiL(k) -WP ( k ) up(k) u P(k) (12)


160

The value of п(convergence coefficient) decides the rate of convergence and

accuracy of estimation. The practical range of convergence coefficient lies in

between 0.1 to 1.0. Three-phase source reference currents corresponding to

positive sequence real component of load current may be computed as:

IV. MATLAB BASED MODEL OF DSTATCOM SYSTEM

Fig. 5 shows the basic simulation model of the DSTATCOM system. The

considered load is a combination of resistance and inductance connected in

series for each phase. The load is star connected 32kVA at 0.8pf The

DSTATCOM model is simulated with above described p-q theory, SRF theory

and Adaline based theory. Figs. 6(a)-(c) show the simulation models for these

theories. The model assembled using mathematical blocks of SIMULINK block-

set. The simulation is carried out in continuous mode at maximum step size of

1*10-6 with odel5s (stiff/NDF) solver

Fig. 5 : MATLAB based model of DSTATCOM System


161

Fig. 6(b) - MATLAB Model of IRP theory

Fig. 6(b) : MATLAB Model of SRF theory

Fig. 6 (c) : MATLAB Model of ADALINE theory


162

V. RESULTS AND DISCUSSION

The performance of DSTATCOM is studied for all three methods of control

techniques. The following observations are made based on these results.

A. Control of DSTATCOM by p-q Theory

Fig. 7 shows the dynamic performance of DSTATCOM using the p-q theory

based current extractor. The considered load is reactive at 0.8 lagging power

factor. The iref is the real part of the load current. The load has been increased

from 16kVA to 32kVA at 0.12sec and unbalanced is introduced at 0.18 sec.

After 0.24 sec the dynamics are shown in reverse sequence. The operation of

DSTATCOM controlled by the p-q theory is shown in Fig. 8. The dynamics and

unbalance conditions are simulated as per the previous case. The delay in

compensation can be seen from source current waveforms. This delay is due to

the low pass filter (LPF) used for filtering power signals. Moreover, p-q theory

uses voltage signals to compute instantaneous active and reactive powers, any

distortion and unbalance in voltage will lead to inaccurate calculation of

reference source currents which should contain only real fundamental frequency

component of load current.

B. Control of DSTATCOM by SRF Theory

Fig. 9 shows the extraction of real fundamental current component of load

current by SRF theory and the performance of the DSTATCOM is controlled by

SRF theory is presented in Fig. 10. The simulation is carried out for similar load

changes and unbalanced conditions as of previous case. The effect of delay due

to LPF used for filtering signals in d-q frame can be seen in extracted reference

currents waveform in Fig. 9. The generation of voltage templates (sine and

cosine) plays an important role in calculation of reference source currents. These

templates are generated using PLL and therefore the tuning of PLL is crucial.


163

The operation of PLL slows and it also imposes some amount of delay in

computation.

C. Control of DSTATCOM by Adaline based Algorithm

Extraction of reference currents by Adaline based current DSTATCOM is

shown in Fig. 12. It can be observed that the DSTATCOM controlled by the

Adaline is able to meet the load changes within one cycle of the sine wave. The

advantage of the Adaline based extractor is that it requires less computation

efforts and therefore the implementation of this technique is much simpler.

Moreover, the inherent linearity of Adaline makes it as a fast technique. The

speed of convergence can be varied by varying the value of η (convergence-

coefficient).

Fig.7 : Reference current extraction using P-Q theory


164

Fig. 8 : Dynamic performance of DSTATCOM controlled by P-Q theory


165

Fig. 9 : Reference current extraction using SRF theory


166

Fig.10 : Dynamic performance of DSTATCOM controlled by SRF theory


167

Fig. 11 : Reference current extraction using ADALINE theory


168

Fig. 12 : Dynamic performance of DSTATCOM controlled by ADALINE

theory

VI. CONCLUSIONS

The mathematical derivation of the p-q and SRF theory has been employed

to demonstrate the behavior of DSTATCOM.

An Adaline based control technique has resulted in considerable improved

performance of DSTATCOM. The Adaline based technique utilizes LMS

algorithm to calculate the weights, and all the calculations are performed online

therefore the algorithm is able to extract reference current components in case of

varying load condition which otherwise is not possible with other neural network


169

based current extraction techniques. MATLAB based results have verified the

effectiveness of these control algorithms.

VII. APPENDIX

Voltage Source converter: DC link capacitor Cdc=1000µF, AC

inductor=4mH, Ripple Filter: C=10µF and R=8Ω, fs=20kHz. Supply voltage

415V, Line frequency-50HZ


170

VIII. REFERENCES

1. Acha, V.G. Agelidis, O. Anaya-Lara and T.J.E. Miller, Power Electronic

Control in Electrical Systems, England, Newnes, 2002.

2. Ghosh and G.Ledwich, Power quality enhancement using custom power

devices, London, Kluwer Academic Publishers, 2002.

3. V.R. Dinavahi, M.R. Iravani and R. Bonert, “Real-time digital simulation

and experimental verification of a D-STATCOM interfaced with a digital

controller,” International Journal of Electrical Power & Energy Systems,

Vol. 26, No. 9, Nov. 2004, pp. 703-713.

4. Akagi, Y. Kanazawa and A. Nabae, “Generalized theory of the

instantaneous reactive power in three-phase circuits,” in Proc. of IEEE

IPEC, 1983, pp. 821-827.

5. D.M. Divan, S. Bhattacharya and B. Banerjee, “Synchronous Frame

Harmonic Isolator using Active Series Filter,” in Proc. of European Power

Electronic Conference, 1991, pp. 3030-3035.

6. G.D. Marques, “A comparison of active power filter control methods in

unbalanced and non-sinusoidal conditions,” in Proc. of IEEE IECON, Vol.

1, 1998, pp. 444 –449.

7. V. Soares, P. Verdelho and G. D. Marques, “Active Power Filter Control

Circuit Based on Instantaneous Active and Reactive Current id-iq Method,”

in Proc. of IEEE PESC, 1997, pp. 106-1101.

8. P. K. Dash, D.P. Swain, A.C. Liew and S. Rahman, “An adaptive linear

combiner for on-line tracking of power system harmonic,” IEEE Trans. on

Power Systems, January 1996, pp.21-25.


171

9. N. Pecharanin, “An application of neural network to harmonic detection in

active filter,” in Proc. of WCCI-ICNN’94, 1994, Vol. 6, No 684, pp. 3756.

10. Pechanranin, H Uitsui and M. Sone, “Harmonic detection by using neural

network,” in Proc. of IEEE International Conference on Neural Network,

1995, Vol. 2, pp. 923-926.

11. B.N. Singh, A. Chandra and K. Al-Haddad, “DSP-based indirect current

controlled STATCOM I. Evaluation of current control techniques,” IEE

Proceedings Electric Power Applications, Vol. 147, no 2, March 2000,

pp.107 - 112.

12. Singh, J. Solanki and V. Verma, “Neural Network Based Control of

Reduced Rating DSTATCOM,” in Porc. of IEEE-INDICON, 2005, pp. 516-

520.

13. Advanced electronics by NED MOHAN

14. Widrow and Mchal A. Lehr, “30 years of adaptive neural networks

perception, madaline and backpropagation,” IEEE Proceedings, Vol. 78,

September 1990, pp. 1415-1442.

15. B. Widrow, J.M. McCool, and M. Ball, “The complex LMS algorithm,”

IEEE Proceedings, Vol. 63, 1975, pp-719-720.

performance analysis of d-statcom ... analysis of d-statcom compensator using control techniques for...

Documents