MATLAB Functions for LPTV Circuit Modeling
MATLAB Functions for LPTV Circuit Modeling
Abstract: The paper presents developed User-Defined The system UDF MAOPCs consists of 17 functions. The
Functions, that are designed for identification of linear results obtained in the process of performing one function can
periodically-time-variable circuits and inverse transformation of be variables for other functions. User can complement the
signals by Fourier and Laplace. Such User-Defined Functions are system UDF MAOPCs by the new functions thereby the
addition to the system UDF MAOPCs, which is designed for existing library will expand by the methods and algorithms of
modeling, multivariate analysis and optimization of linear
computational experiments which are needed for research of
periodically-time variable circuits in the frequency and time
domains. The paper also presents computer experiments of the LPTV circuits.
usage of these developed User-Defined Functions. Currently in the system UDF MAOPCs are absent functions
Keywords: the system UDF MAOPCs, linear periodically- that solve such important tasks as: a) the inverse Fourier
time-variable circuits, the inverse Fourier transform, the inverse transform and the inverse Laplace transform of signals that are
Laplace transform, identification, the frequency symbolic defined in the frequency domain; b) the identification of LPTV
method. circuits. In connection with this, the paper is dedicated to
creation and description of these functions, marked as
I. INTRODUCTION «InverseLaplaceTransform» and «Identification»,
Devices, that are modeled by linear periodically-time- respectively. The paper also presents computer experiments
variable (LPTV) circuits, are widely used in modern electrical with single- and double-circuit parametric amplifiers,
and radio systems, computer technology, systems of automatic conducted in the system UDF MAOPCs computer, which
control and communications. Therefore, the task of modeling confirmed the adequacy of the created functions.
LPTV circuits is actual [1,2,3]. II. MAIN PART
With the advent of powerful software of symbolic
mathematics, there have started to appear the specialized The inverse Laplace transform and the inverse Fourier
software products that implement symbolic methods in the transform of signals that are defined in the frequency domain.
aided design, in particular, in the aided design of electrical and In the system UDF MAOPCs we introduce the function
radio engineering devices and systems. One of such products «InverseLaplaceTransform», which is designed to create time
is the system UDF MAOPCs (User-Defined Functions dependence y (t ) of the output signal (current or voltage) of
Multivariate Analysis and Optimization of the Parametric the circuit based on its image Y ( s, t ) in the frequency domain.
Circuits) [4,5], which was created in MATLAB environment
and is based on the frequency symbolic (FS) method [6] and Obviously, before performing this function in the program of
also is designed for modeling, multivariate analysis and research of circuit, we must perform: a) the function «laplace»
optimization of LPTV circuits in the frequency and time for transformation of the input signal x(t ) to the frequency
domains [7]. There are implemented functions in the system domain X ( s ) ; b) the function «TrFunc», which forms
UDF MAOPCs that allow to solve the following tasks:
parametric transfer function W ( s, t ) that relates the input
determination of parametric transfer functions of LPTV
circuit in the frequency domain; X (s) and output Y ( s, t ) signal of circuit; c) the function
determination of instantaneous values of voltage and «OutVar», which forms output signal Y ( s, t ) as a
current in the branches of LPTV circuit;
multiplication Y ( s, t ) W ( s, t ) X ( s ) .
assessment of the stability of LPTV circuit;
determination of the instantaneous power in the branches The parametric transfer function W ( s, t ) by the FS-method
of the LPTV circuit and the transfer coefficients of average has the form of polynomial Wˆ ( s, t ) of Fourier in the complex
power from input circuit to it's output;
form:
determination of sensitivity coefficients; r
ªW i ( s ) exp( ji: t ) º
tolerance analysis; Wˆ ( s, t ) Wr0 ( s ) ¦ « », (1)
optimization; i 1 ¬ W i ( s ) exp( ji: t ) ¼
2D and 3D visualization of the results of research of where Wr0 ( s ), W i ( s ), W i ( s ) - are the time-independent t
LPTV circuits. fractional-rational functions of a complex variable s; r- is the
Authorized licensed use limited to: Peter the Great St. Petersburg Polytechnic Univ. Downloaded on March 29,2022 at [Link] UTC from IEEE Xplore. Restrictions apply.
161
number of harmonics in the approximation; : - is the greatest 3. In algebraic expression (5) the number of unknown
common divisor of the frequencies of periodic changes of coefficients is (n m 1) , therefore, for their determination
parameter of parametric elements of circuit under the influence we write down expression (5) for different (n m 1) values
of the parameters of the pumping signal; j 2 1 . [ 2, ..., [ n m 1 , as a result it is forming a system of linear
Function «InverseLaplaceTransform» calculates the signal algebraic equations (SLAE) order (n m 1) :
y (t ) using the inverse Fourier transform for the steady-state
an (t ) y ( n ) (t , [1 ) an 1 (t ) y ( n 1) (t , [1 ) ... y (t , [1 )
mode
f bm (t ) x ( m ) (t [1 ) bm 1 (t ) x ( m 1) (t [1 ) ...
1
y (t ) ³ W ( j Z , t ) X ( j Z ) e j Z t d Z (2) b0 (t ) x(t [1 ),
2 S f
or the inverse Laplace transform for the transition mode ............................................................................ (6)
1
c j f
an (t ) y (t , [ m n 1 ) an 1 (t ) y
(n) ( n 1)
(t , [ m n 1 ) ...
2 S j c ³j f
y (t ) W ( s, t ) X ( s )e s t ds. (3)
y (t , [ m n 1 ) bm (t ) x (t [ m n 1 ) bm 1 (t )
(n)
Authorized licensed use limited to: Peter the Great St. Petersburg Polytechnic Univ. Downloaded on March 29,2022 at [Link] UTC from IEEE Xplore. Restrictions apply.
162
the identified coefficients with coefficients that were 3. By the function «OutVar» we form the image of output
determined by the structure and parameters of the amplifier. signal of the amplifier
U (s,t,[ ) W (s,t ) I (s,[ ) P1 (s,t ) P2 (s,[ ) / Q1 (s ) Q2 (s ) , (11)
where P1 (s,t ) - numerator of function W (s,t ) ; P2 (s,t ) -
numerator of image of the input signal; Q1 (s ) - denominator
of function W (s,t ) ; Q2 (s ) - denominator of image of the input
signal.
4. By the function «InverseLaplaceTransform» we form
(n m 1) output signals u (t,[i ) in the time domain for
[ [1 , [ 2, ..., [ n m 1 , repeatedly using the expression [8]:
m
P1 (sm ,t ) P2 (sm ,[i )
u (t,[i ) ¦d e sm t , (12)
Q1 (sm ) Q2 (sm )
m 1
ds
where sm - roots of equations Q1 (s ) 0 and Q2 (s ) 0 .
The instantaneous values of the output voltage were
calculated by the expression (12) for the
[ 0.004 ȝs, are
presented in Table 1. In Table 1 and Table 2 -Table5 the results
Fig. 2 The algorithm of function «Identification» are presented with four decimal places, although the
calculations are made with bitness of fifteen decimal places.
TABLE 1. THE INSTANTANEOUS VALUES OF THE OUTPUT VOLTAGE OF
AMPLIFIER FROM FIG.3 FOR [ 0.004 ȂS
t u(t), V u(t), V
Fig.3. Single-circuit parametric amplifier
ȝs by MATLAB by Micro-Cap7.0 G ,%
c(t ) c0 (1 mc cos(:t Ic )), c0 10 1012 ɎY mS mc
L 0, 2533 106 H , : 4 S 108 rad / s, Z 2S 108 rad / s 1.000 -1.5012 -1.5012 0.112e-3
i (t [ ) 0.001 cos(Z (t [ )) 0.004 cos(2 Z (t [ )) 1.001 -0.5505 -0.5505 0.148e-3
The progress of the experiment. Step 1. Application of From Table 1 it follows that instantaneous values of the
functions «InverseLaplaceTransform». output voltage of amplifier that were formed by the system
UDF MAOPCs, coincide with the values that were obtained
The output signal u (t , [ ) we define by functions «TrFunc»,
using the Micro-Cap7.0, with the relative error of 0.197e-3% -
«laplace», «OutVar» and «InverseLaplaceTransform» in the 0.638e-3%.
time domain. The procedure of formation u (t , [ ) of amplifier Step 2 The application of the function «Identification».
from Figure 3 in the system UDF MAOPCs is as follows: We consider that the coefficients of equation (8) are
1. By the function «TrFunc» we form: unknown and variables of this equation are known. By the
- differential equation wich is describing a given circuit function «Identification» we define derivatives of variables of
relatively to the input current i (t [ ) and output voltage equation (8) and we identify it in the form of the equation of
u (t , [ ) , [6]: the second order:
a2 (t ) u '' (t , [ ) a1 (t ) u ' (t , [ ) a0 (t ) u (t , [ )
§ 1· (13)
c(t )u cc(t , [ ) 2cc(t ) Y u c(t , [ ) ¨ ɫcc(W ) ¸ X (W , [ ) L c(W [ ); (8)
© L ¹ b2 (t )i '' (t - [ ) i ' (t [ ) b0 (t )i (t [ ).
- differential equation of L.A. Zadeh [6]: In this algebraic expression (13) the number of unknown
L c(t ) W cc(s,t ) 2 L c(t )s 2cc(t ) L L Y W c(s,t ) coefficients are (n m 1) =5, so the determination of the
(9) expression (13) we write down for 5 values [ [1 , [ 2 , [3 , [ 4 , [5
L c(t )s 2 (2cc(t ) L L Y )s (ccc(t ) L 1) W (s,t ) Ls;
resulting in forming SLAR of fifth order. Solution of SLAR at
- the parametric transfer function W (s,t ) for the r =6 selected time points gives the value of the identified
harmonic components in the form of (1). coefficients (13) at the same time-points.
2. By the standard function «laplace» for symbolic variables The identified coefficients of equation (13) by the function
s,[ we define the image of input signal i (t [ ) as: «Identification» at selected time points and calculated exact
values of the same coefficients at the same time points by
103 s cos (2S 108 [ ) 2S 105 sin (2S 108 [ )
I (s,[ ) equation (8) are presented in Table 2.
s 2 4 S 2 1016
0.004(s cos (4S 108 [ ) 4S 108 sin (4S 108 [ )) TABLE 2. THE INSTANTANEOUS VALUES OF COEFFICIENTS a0 (t ) ,
. (10)
s 2 16 S 2 1016 a1 (t ) , a2 (t ) , b0 (t ) , b2 (t ) OF EQUATION (13) AND EQUATION (8)
Authorized licensed use limited to: Peter the Great St. Petersburg Polytechnic Univ. Downloaded on March 29,2022 at [Link] UTC from IEEE Xplore. Restrictions apply.
163
Authorized licensed use limited to: Peter the Great St. Petersburg Polytechnic Univ. Downloaded on March 29,2022 at [Link] UTC from IEEE Xplore. Restrictions apply.
164
TABLE 4. THE INSTANTANEOUS VALUES OF THE COEFFICIENT a4 (t ) , (Table. 4) do not significantly affected on accuracy of the
a3 (t ) , a2 (t ) , a1 (t ) , a0 (t ) , b3 (t ) , b2 ( t ) , b0 ( t ) OF EQUATION
calculated values of the voltage uc (t ) of amplifier from the
(15) AND EQUATION WHICH IS DESCRIBING A GIVEN CIRCUIT Table. 5.
RELATIVELY TO THE INPUT CURRENT AND VOLTAGE III. CONCLUSION
t,ȝs a4 (t ) , identified a4 (t ) ,exact G% 1. The results of computer experiments that were performed
with single and double circuit parametric amplifiers convince
100.0 ɟ-28 ɟ-28 0.2373 us in adequacy of the of functions «InverseLaplaceTransform»
t,ȝs a3 (t ) , identified a3 (t ) , exact G% and «Identification» that were developed in the system UDF
100.0 -ɟ-20 -ɟ-20 0.3574 MAOPCs.
2. Considered in this paper functions
t,ȝs a2 (t ) , identified a2 (t ) , exact G%
«InverseLaplaceTransform» and «Identification» expand the
100.0 0.8733e-10 0.8729e-10 0.2459 possibilities of the system UDF MAOPCs, that is based on the
t,ȝs a1 (t ) , identified a1 (t ) , exact G% frequency symbolic method in the tasks of investigations of
LPTV circuits in the time domain.
100.0 0.0028 0.0028 0.3044 3. The function «Identification» allows us to create models
t,ȝs a0 (t ) , identified a0 (t ) , exact G% of parametric devices like "the black box" that can be useful in
100.0 ɟ ɟ 0.6282
industrial manufacturing of parametric devices or when using
such devices in electrotechnical or radiotechnical systems.
t,ȝs b3 (t ) , identified b3 (t ) , exact G% 4. The developed function «InverseLaplaceTransform» of
100.0 6.3245e-19 6.3325e-19 0.1292e-2 the system UDF MAOPCs, for the first time allows to obtain
t,ȝs b2 ( t ) ,identified b2 ( t ) , exact G% time dependence of signals by the inverse Fourier transform
and the inverse Laplace transform on engineering level at the
100.0 9.3124e-13 9.3126e-13 0.4969
calculations of practical parametric circuits.
t,ȝs b0 ( t ) ,identified b0 ( t ) , exact G% 5. The developed functions «InverseLaplaceTransform»
-0.2830e4 - and «Identification» in the environment of the system UDF
100 0
MAOPCs can successfully solve tasks of statistical researches
From the Table 4 it follow that identified values and exact and optimization of LPTV circuits in the time domain.
values of coefficients coincide with the relative error 0.1292e-
2% - 1.2502%. REFERENCES
The identified coefficients ai (t ) and b j (t ) from Table.4 by [1] Tholen, E.A. et al. Nonlinearities and parametric
the function «Identification» we approximate by Fourier amplification in superconducting coplanar waveguide
polynomials that at selected value k 19 contains q 9 resonanors. Appl. Phys. Lett.90, 253509(2007).
harmonic components. [2] Byeong Ho Eom, Peter K. Day, Henry G. LeDuc and
For the determination of the adequacy of the approximated Jonas Zmuidzinas. A wideband, low-noise
coefficients in Table 5 for input signal i (t ) 0.001 cos(Z t ) superconducting amplifier with high dynamic range.
Nature Physics, Vol.8, August 2012.
are shown value of the voltage uc (t ) , are formed by the
[3] Vanassche P., Gielen G. and Sansen W. Systematic
identified equation and calculated by the program of numerical Modeling and Analysis of Telecom Frontends and their
analysis Micro-Cap7.0 by the exact equation which is Building Blocks Springer; 2005, ɪ
describing a given circuit relatively to the input current
[4] Yu. Shapovalov, B. Mandziy and [Link] «The system
i (t [ ) and voltage uc (t , [ ) . functions MAOPCs for analysis and optimization of linear
TABLE 5. THE INSTANTANEOUS VALUES OF THE VOLTAGE uc (t ) OF
periodically time-variable circuits based on the frequency
symbolic method», Przeglad Elektrotechniczny, vol.91,
AMPLIFIER FROM FIG.4
no 7, pp. 39-42, 2014.
t,ȝs uc (t ) , V uc (t ) , V [5] MATLAB. Product Documentation [Online]. Available:
by MATLAB by Micro-Cap7.0 G ,% [Link]
100.0 1.9414 1.9497 2.5760e-2 [6] Shapovalov Yu. Symbolic analysis of linear electrical
circuits in the frequency domain. Fixed and variable
100.1 1.8615 1.8606 2.1498e-2
parameters. Lviv, Lviv Polytechnic National University
From Table 5 it follows: publication, 2014, 324 p.
a) the instantaneous values of the output voltage of the [7] Yu. Shapovalov, B. Mandziy, and S. Mankovsky, «The
amplifier that were calculated in the system UDF MAOPCs, peculiarities of analysis of linear parametric circuit
coincide with the values that were obtained using the program performed by frequency-symbolic method», Przeglad
Micro-Cap 7.0 with relative error 2.1498e-2% –6.0945e-2%; Elektrotechniczny, vol.86, no 1, pp. 158-160, 2010.
b) unequal zero identified coefficients b0 (t ) (Table. 4), and [8] G. Korn and T. Korn. Mathematical Handbook for
some enough inaccurate identified values of coefficient a1 (t ) Scientists and Engineers, - M .: Science, 1974, 832 p.
Authorized licensed use limited to: Peter the Great St. Petersburg Polytechnic Univ. Downloaded on March 29,2022 at [Link] UTC from IEEE Xplore. Restrictions apply.