Transient Analysis in Electrical Circuits using

Z-transformation

Branislav Dobrucký, Pavol Štefanec, Roman Koňarik Oleg V. Chernoyarov

Faculty of Electrical Engineering Department of Radio Engineering Devices
University of Žilina and Antenna Systems
Žilina, Slovak Republic NRU - Moscow Power Energetic Institute
branislav.dobrucky @fel.uniza.sk Moscow, Russia
O_v_č@fel.uniza.sk

Abstract—Paper deals with mathematical modelling of impulse Applying Z-transform and inverse Z-transform [7]-[10] we
waveforms and impulse switching functions used in electrical can determinate K for any time instant:
engineering. Using z-transformation the impulse switching
functions can be obtained for any discrete time instant as impulse   _ 1   \ 1 bc
sequences of exciting functions. Applying such an impulse  bc b)c
sequence to electrical circuit the response of state variables can be   _ 
investigated using inverse z-transformation. Theoretical derived  1  ] 1
wave-forms are verified by LT-Spice circuit simulator and where   as sum of geometrical series is equal
compared to those of simulation worked-out results.

  _  b*c
Keywords—
impulse switching function; Z-transform; inverse Z- ]&
transform; modelling and simulation; steady state operation;
Then
dynamical state model.
 & 
I.
TRANSIENT ANALYSIS USING Z-TRANSFORM   _ 1 b+c
  ] 1  ] &
Many papers are dealing with modeling of dynamical system
using complete and expensive programing environment. Pro- With roots of denominator
posed simulation model is much cheaper and suitable option for : _ 1 
; _ &
performance analysis under various conditions as in the real
system [1]-[2]. Using the Cauchy method of residues [12]-[14]
G
A.
Z-transform analysis in DC circuits
Let’s take a sinusoidal voltage function K _   ] L    KC;
Q2QX
LD;
  _  _
 
 :K ;K
connected to the R-L circuit. The current flowing through the K _ 1 \ _
 : ] ; ; ] : b,c
circuit is [3]-[5]
K
6 V V  1 &
8 9 _ & ] 7 CW \ 4: 7 CW & _ 1 \ _
5  1 ] & & ] 1
Main advantage of analytical solution is possibility of 
_ & ] 1 K 
calculation of the value in any time instant inserting  into 
equation   _   Numerical solution does not make possible
Let 1 is "&%% then
that benefit:

bc & & K _ & ] %..*K  b-c
_]   \ b'c 
 " "
Or, when step of sequence we set to "
1 1 
KB; _ & ]  \ _ 
" K "  M _ & ] %(++M  -
( 

_  1 K \  1  where T and T are values for  _ &%%.



978-1-4673-8698-2/16/$31.00 ©2016 IEEE 131

 If we take 1 _ (+%  $1 _  !&-% and

 $1 _
 !&-% then
F &   !&-%
  _ 1 <
  ] 1  ] '
 !&-% \ &
b&*c
F   !&-%
_ 1
  ] :  ] ;  ] <
where
: _ 1 
;< _ %...- ^ %%().
Fig. 1 Waveform of current response with 1-step
and, if " _ '
Steady state value determination
: _ & ] & &-% _ &,. &-% _ %..))
 & 
  _ 1 b.c Using the Cauchy method of residues (7)
  ] 1  ] &
F ! &
K2U _   ] &   = K _ 1  K
Q2;B
EM
QD:
 &-% : ] ; : ] < :
b&%c &
 &  \ K
_ 1 _  ; ] : ; ] < ;
 & ] 1  &
\ K _
The same result as using analytical solution. < ] : < ] ; <
(16)
B.
Z-transform analysis in AC circuits F 1 &
_ 
Let’s take a sinusoidal voltage function  ' 1 < ] '%%&'1 \ &
  _ F 
b$c !
  K \  1 ] CJR JKR
&-% 1
connected to the R-L circuit. The current flowing through the
] ] 1 ] JR CJKR
circuit is [3]-[5]
1H 1H
F where 1 _ & ] ; 1 _ and if 1 _ ', so
  _  $ ] # \  #  CPT T T
 b&&c 1 _0.9944, and _ '%%&'/ el.
\ : CPT
After calculation using [12]-[14] we get
where 
#
" are known, and depend on parameters if the
circuit. F 1
K _

 1 < ] '
 1 \ &
Classical numerical solution of ODE:
! b&,c
bc & & F    K \
    ] 1  
_]   \  $  b&'c &-% 1
 " "  ]  
 
Using discretization by Euler’s method
Another way is to use new method with help from [8] and
1 1 F applied in [19]
KB; _ &] K \  _






" "  K
&( :
F F
_  1 K \  1  K _  1 K K \  1 1 L KCbLB;c b&-c
 K 
LDKC;
where the K _  $1  .
Applying Z-transform:
F
  _ 1   \ 1 bc

b&)c
F bc
  _ 
 1  ] 1
where
  $1
  _  b&)c Fig. 2 Waveform of current response with 1-step
 < ] '
 $1  \ &

132
 Worked-out results are practically identical with those which Applying inverse transform (see Appendix)
obtaining by analytic solutions.
& ! ! !
  _
  \ '   \
II.
TRANSIENT ANALYSIS IN CONVERTER SYSTEMS USING Z- ( ' + + b'%c
! !
TRANSFORM \ ]& N '   \
+ +
Converter system output voltages and currents are impulse S
functions especially in case of voltage sourced inverters (VSI) where  is an index of sequence with step .
@
with pulse-width-modulation (PWM), Fig. 3.
It being understood that sequence period &' may not be
consistent with the time period of the waveform  (preferably,
if it is less).
By substitution
&'
 _  1 b'&c

We obtain
  _
& &' !
Fig. 3 Three-pulse modulated waveform _
 b 1c

















(  '
&' ! ! b''c
By modeling using we can create impulse switching function \ ' b 1c \
 + +
@ ' > < INP
;<
1HK &' ! !
=OZ[Y  _ \ \ _ \ ]& H ' b 1c \
@ \ & @ \ & @ \ &  + +
Dynamical state model of the systems include exciting
b&.c functions   as input vector. The models can be expressed in
 <  > \ ' < \ & a discrete form as:
_

@ \ & KB; _ 1 K \ 1 K b'(c
where roots of polynomial of the denomina-tor are or
S S ?S NB; _ H;< N \ H;< N b')c

;< _ ^I < 

=> _ ^I @ 
?@ _ ^I @
placed on boundary of stability in unit circle as in Fig. 4. respectively, where  is order of computation step and  as
above.
bc & &
_]   \ 
bc b'*c
 " "
1 1 
KB; _ & ] K \  _






" "  K
'+

_  1 K \  1  K

where K _    For calculation K following Eq. can be
used [..]
Fig. 4 Pole placement of denominator
:
Its projection is shown in Fig. 5. K F
K _  1 K \  1 1 L KCbLC;c b',c

LDKC;
and taking
  1 &
"_ 
1 _ 
_ _ %%%*+
' (+% " &-%
Discrete form of state space model of the investigated system
with the step of impulse switching function can be obtained
directly from the impulse switching functions [5], [11]-[15]
Now, we can draw the function for  3` %
0 a thus it can
be solved dynamical state of the system given by Eqs. (26), (27)
Fig. 5 Impulse switching function of the three-pulse waveform. continuously with the 1 com-putation step, Fig. 6.

133
Fig. 6 Waveform of sequence Kc Fig. 9 Waveform of voltage and current sequences H;<c and H;<c

Verifying by LT Spice circuit simulator are presented in Fig.

10 and Fig. 11.
A1
GS1 GS1 D1 GS4 D2
a
>e V2
b

S1

S4
V1 V4 A4
GS2
0.5 R1 L1
A2
0.9 a
~~ V3
b
V5 A5 GS3 GS2
GS3 0.5 D4 D3

0.9 a

S3

S2
<e A3
b
GS4
PWL file=sig2X.txt

Fig. 7 Waveform of sequence Kc .tran 0 20m 0 1e-6 uic

.model MYSW1 SW(Ron=0.001 Roff=1Meg Vt=0.8 Vh=-0.01)

Fig. 10 Generating of three-pulse exciting voltage waveform

The worked-out results are shown in Fig. 11-12.

Fig. 8 Verification in Matlab environment

If taking into account voltage N and

:
=:
H;< _ 1 


HA _ 1 1 L
LD=:C; Fig. 11 Worked-out original record
Then V(N006,N007) I(L1)
1.2V 1200mA
N 1.0V 1000mA
N _  H : \  H 0.8V 800mA
;< ;<
0.6V 600mA
: 0.4V 400mA
L
 H NC LC; _ %-)+& N : \ 0.2V 200mA

;< b'-c 0.0V 0mA

LDNC; -0.2V -200mA

-0.4V -400mA
:
& L -0.6V -600mA

\ H NCbLC;c 
-0.8V -800mA

&-% ;< -1.0V -1000mA
LDNC; -1.2V -1200mA
0.000ms 3.332ms 6.664ms 9.996ms 13.328ms 16.660ms 19.992ms

Fig. 12 Worked-out results – from extracted data

By similar way as above can be investigated higher order

systems. Next, there is show .. second order R-L-C electrical
circuit under supply by 3-pulse voltage, Fig. 13, [16]-[18].

134
 & S S S S S S S ?S
_ & JN < \ & CJN < \ 'CJ = JN @ \ 'J = CJN @ \ 'J = JN @
+
S ?S
\ 'CJ = CJN @ _

! ! !

  \ (  \
 
& ' + +
_ ?S ?S ?S ?S _
(CIN @ ] IN @ CIN @ \ IN @
] ( \
' '
! ! !
&
  ' \ (  + \
  +
Fig. 13 Worked-out results of second order system _ ! ! _
( \ ]& N ( 
 
+ +
At those simulations have been used:
& ! ! ! ! !
@ ' > < _ d
  \ '  \ \ ]& N '  \ e
  _
\ @ \ @ ( ' + + + +
@
\&  \&  \&
ACKNOWLEDGMENT
! ! ! !
N _   \ \ ]& N   \ The paper was supported from Slovak Grant Agency VEGA by
+ + + +
the grant No. 1/0928/15 and the Russian Science Foundation by
;
the grant No. 14-49-00079.
< NB; _ H ;< N \ H ;< N
in incremental form, or in absolute form
REFERENCES
:
; N L
< _ H : \ H H NCbLC;c [1]
J. Dudrik, P. Bauer: New methods in teaching power electronics and
N ;< ;< ;<
LDNC; devices. In: International journal engineering education IJEE, Vol. 24, No.
5, 2008, ISSN 0949-149x, pp.1040-1048.
By this way it is possible to investigate, of course, also the
[2]
M. Galád, J. Šedo, P. Špánik: Developing of Stand-Alone Power System
systems of higher order > 2. Simulation Model. In: International Review of Automatic Control –
IREACO, Vol. 7, No. 5, 2014, pp. 500-505.
III.
CONCLUSION [3]
J. G. Aramovich, G. L. Lunts, L. C. Elsgolts: Functions of Complex
Comparing Fig. 6 and Fig. 7 to Fig. 8 and Fig. 11, Variable, Operator Calculus, Stability Theory (in Russian – translated
respectively, it is clear that verified waveforms exactly confirm into Slovak), Alfa, Bratislava, 1973.
the calculated and simulated waveforms. Main qualitative [4]
E. Ušák: Transient phenomena solution using integral transforms (in
Slovak). Textbook, STU Bratislava, 2011
difference between method given in the paper and classical
[5]
R.J. Beerends, H.G. Morsche, J.C. Berg, E.M. Vrie: Fourier and Laplace
numerical solution is that described method makes possible to Transforms. Publisher: Cambridge University Press, 2003.
obtain the instantaneous values of investigated quantities at any
[6]
T.J. Takeuchi: Method of Φ-Function to Analyses of SCR Rectifier
time instant (k.
1) thus this new method practically equals to Circuit. In: Journal of IEEJ, Japan, Vol. 83-10, No. 901, 1963.
analytical solution. Behaviour of the system can be estimated [7]
Ja.Z. Cypkin: Theory of linear impulse systems (in Russian). GIF-ML
from pole assessment of the polynomial of the transfer function Publisher, Moscow, 1963.
denominator similarly as analytical solution or operator calculus [8]
V. Strejc: State-space theory of linear discrete control (in Czech).
(Laplace transform). If one can to determine value of Academia Publisher, Praha, 1978, ISBN 509-21-857, Chap. 2.
investigated quantity at time instant 1 analytically then [9]
R. Vích: Z-transformation and some of its utilization (in Czech). SNTL
waveforms at any time instant (k.
1) will be entirely consistent Mathematical Seminar, Prague, 1983.
with analytical solution. [10]
V.P. Shipillo: Determination of original function at discrete electrical
circuits and systems (in Russian). „Elektritchestvo“ Journal, 1981, no. 1,
pp. 52-55.
[11]
B. Dobrucký, O.V. Chernoyarov, M. Marčoková: Computation of the
APPENDIX total harmonic distortion of impulse system quantities using infinite
& series. Proceedings of 14th Conference on Applied Mathematics, STU
N _
 Bratislava, Feb 2015, pp. 213 – 220.
+ [12]
I.V. Blagouchine, E. Moreau: Analytic method for the computation of the
S S total harmonic distortion by the Cauchy method of residues. IEEE
& \ CJS \ CJ<S JN < \ & \ JS \ J<S CJN < Transactions on Communications, Vol. 59, No. 9, Sept. 2011: s. 2478-
S <S S S <S S 2491.
\ & \ CJ = \ CJ = JN @ \ & \ J = \ J = CJN @ [13]
J. Moravčík: Mathematical Analysis 3 (in Slovak), Textbook, Alfa
S <S ?S S <S ?S Publishing House, Bratislava, 1992, ISBN 80-05-01044-3.
\ & \ J = \ J = JN @ \ & \ CJ = \ CJ = CJN @ [14]
J. Hrivňák, N. Spurná: Linear digital control circuits: Examples and tasks
(in Slovak). ALFA Publisher, Bratislava, 1983.
[15]
T.K. Moon, W.C. Stirling: Mathematical methods and algorithms for
signal processing. Prentice Hall, New York, 2000, Chap. 3.

135
[16]
B. Dobrucký, M. Pokorný, M. Beňová: On Impulse Switching Functions [18]
B. Dobrucký, P. Štefanec, M. Beňová, O. V. Chernoyarov, M. Pokorný:
of Inverters as an Orthogonal System. Journal of Mathematics and System Analysis of the 2nd Order System with Impulse Exciting Functions in Z-
Science (USA), ISSN 2159-5291, Vol. 4 (2014), No. 12, pp. 743-749. Domain - submitted to Hindawi Journal: Mathematics problems in
[17]
B. Dobrucký, M. Pokorný, M. Beňová, M. A. R. Abdamula: Modeling of Engineering for publications in Nov. 2015.
power converters using Z-transform. Communication – Scientific Letters
of the University of Zilina, 2013, vol. 15, no. 3, pp. 43-47.

136