Vasyl TCHABAN1,2, Dmytro PELESHKO2
University of Rzeszow (1), Lviv Polytechnic National University (2)
The mathematical model of inverter
Streszczenie. Zapropono model matematyczny nowej generacji jednofazowego inwertora. Transformator jest opisany przez równania potencjału
wektorowego kwazi-stacjonarnego pola elektromagnetycznego. Operacje sterowania tyrystorów opisane są przez wielkości logiczne, które
odzwierciedlają rzeczywisty proces fizyczny. Kondensator jest opisany przez zwyczajne równania różniczkowe. Przedstawiono rezultaty symulacji
komputerowej przebiegów przejściowych. (Model matematyczny przekształtnika)
Abstract. The new generation mathematical model of single-phase inverter is proposed. The iron-clad transformer is described by equation of quasi
stationary electromagnetic field vector potential. The adapter operation of thyristors is described by logical values which reflect the real physical
process. Capacitor is described by ordinary differential equation. The coil resistance is changed for balance resistor. There are shown the results of
computer simulation of model device transient processes.
Slowa kluczowe: model matematyczny, konwertor mostkowy, elektromagnetyczne pole, cieplne pole.
Keywords: single-phase inverter, electromagnetic field, iron-clad transformer, logical variable, C-filter, RL-load.
Introduction where is vector of current density; N, are matrixes of
Modern mathematical theory of high precision electronic reluctivities and conductivities; is the Hamilton’s operator.
systems indispensable run into the problem of the most The module of current density vector we calculate as
modern mathematical models elaboration of devices of (3) k = i k /S k, k = 1, 2
converter technics. The analysis of the problem show that
such mathematical models must be make with employment where Sk is square of cross-section of k-th winding conduc-
of electromagnetic field theory methods only because the tor.
i2
electromagnetic circuit theory methods reach the limit of
their resources completely. In the paper is proposed the i1
field mathematical model of single-phase semi-conductor Lwr
Rb
thyristor for the first time. The inverter is the most
uC uR
widespread electric device of electronic systems.
The mathematical model of inverter Rwr
iC uL
The main elements of inverter are transformer, semi-
conductor thyristor bridge, filter of course, electric voltage
source and load. Its scheme is shown on fig.1 The choking
coil is changed for balance resistor Rb. The electromagnetic
process in iron-clad transformer is described by quasi Fig.1. The scheme of single-phase bridge rectifier
stationary electromagnetic field equations. The semi-con- From the system of equations (1)-(3) we are received
ductor thyristor bridge is circumscribed by logical values such computation equation too [2]
which reflect the real physical process. Condenser is cir-
A A
cumscribed by ordinary electric circuit differential equation. (4) n 0 dl A n 0 u t
The main idea of this articles is concluded in order to t l t l
device duty with condition of thyristor ideal commutation to where n0 is normal vector. This equation is used for given
circumscribe by special control of transformer elec- voltage regime.
tromagnetic field so as to recreate real device duty. In The equations (4) circumscribes magnetization winding
practice we have inverse situation: the time changed non- zones.
symmetry that is created by thyristor duty generates strong
determine behavior of electromagnetic filed in time. y
Therefore we go to backdirection. The field control we
realize by use of control logical variable. As will be shown
later proposed solution gives perfect results: we control the
electromagnetic filed - recreate the perfect device duty. Fe
The perfection of mathematical model of device on the
whole depends on perfection of transformer mathematical
model. Magnet conductor and electric windings are the
base of transformer construction. The cross-section of quar- Cu Cu
ter of transformer body are shown on fig.2.
The equation of electric windings we write as
A z
(1) l t dl uk rk ik , k 1,..., n
k Fig.2. The calculation zone of quarter cross-section iron-clad
where A is vector potential of electromagnetic field; u, i, r transformer
are voltages, current and resistance of winding; l is length of
winding wire; t is time. The integral in left part of equation The equation (2) covers following zones of spatial inte-
(1) is time derivative of full linkage of k-th winding. gration: laminated magnet conductor and air. It changes in
The equation of quasi stationary electromagnetic field each zones. Having the conditions
we use in the following form [1, 2] (5) A = x0A; = x0
A in the zone of laminated magnet conductor section the
(2) 1 A equation (2) transforms into time algebraic equation
t
PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 88 NR 6/2012 251
A A H f Bf
(6) y z 0 (11) f
z z y y Bf
where y, z are y- and z-direction reluctivities of medium. In the air zone the equation (2) will be more simple
Having taken the conditions (5) in the zone of windings
2A 2A
equation (4) assumes new form (12) 0 2 2 0
A c A 2 A 2 A c z y
(8)
t lk x 0 t
dl 2 2 uk
x y z lk
where is reluctivity of medium.
lk Providing the condition (1) and (9) primary current
where x is equivalent conductivity of k-th winding medium; equation we write as
0 is reluctivity of dielectric medium; c is constant coeffici- A
uC x 0 dl
ent. The integral in the left part of (5) is taken over the l1
t
trajectory on the surface of winding conductors. For c =1 the (13) i1
equation (6) describes electromagnetic process in massive r1
conductor. Eddy current need depends if winding are where uC is capacitor voltage.
constructed from thin conductors. This can make easily by According to positive directions of current and voltage
adaption c . capacitor current equation will be
60.0 u uC
(14) iC i1
Rwr
40.0
where u is input voltage; iC is capacitor current; Rb is
balance resistance; is logical variable which can be
20.0 calculated from
if hT t hT T / 2;
u1 , u2 , V
1,
0.0 (15)
1, if hT T / 2 t (h 1)T ,
where T - period; h = 1, 2,... .
-20.0
The capacitor ordinary differential equation assumes
form
-40.0
duC 1 u uC
i1
0.00 0.02 0.04 0.06 0.08
t, s (16)
Fig.3 The time function of primary (1) and secondary (2) voltages of dt C Rwr
transformer where C is condensance of condeser.
In the practical spatial discretization (8) we substitude According to the Kirkhof’s law we receive the differential
the integral for the following sum equation of inductance coil
A
(9) l t
x 0
A
dl = 4 d 2
q2
Am (17) Lwr
di2
dt
( Rwr r2 )i2 x 0 t
dl
m 1 t l2
where d is thickness of magnet conductor. where Rwr is resistance; Lwr = Lwr(i2) is differential
0.10
inductance
The analysis of transient processes of single-phase
inverter is interconnected with integration of differential
0.05
equations (6), (8), (12) of transformer, (17) of inductance
coil and capacitor (16).
0.00
The spatial discretization of partial derivatives has been
i 1 , i2 , A
realized by the finite difference method
-0.05
dy y x x y x x
(18)
dx 2 x
-0.10
d2y y x x 2 y x y x x
-0.15
0.00 0.02 0.04 0.06 0.08
dx 2
x 2
t, s where y is function; x is spatial step of discretization.
Fig.4. The time function of primary (1) and secondary (2) 10.0
transformer currents
5.0
The laminated magnet conductor has been changed by
continuous anizotropic ferromagnetic medium. The
magnetization characteristic of this medium H = H(B) we re- 0.0
u L , uR , V
ceive from recalculation of the ferromagnetic characteristic
Hf = Hf (Bf) by axes x and y [1] -5.0
Bf d f d0
(10) B ; -10.0
d f (d 0 f ( B )) / 0
where d0, df are thicknesses of ferromagnetic sheet and air -15.0
interval; f(Bf) is static reluctivity of ferromagnetic that is 0.00 0.02 0.04
t, s
0.06 0.08
calculated from ferromagnetic magnetization characteristic Fig.5. The time function of inductance coil (1) and resistor (2)
Hf = Hf (Bf) voltages
252 PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 88 NR 6/2012
0.80
The time discretization has been realized by the implicit
method
y(t +t) = y(t) + t f(y, t +t)
0.40
(19)
where f(y, t) is rigth part of space discretizated differential
i1 , i2 , A
0.00
equations.
The simultaneous algebraic equation system
(20) (B + C)x = b -0.40
where B is bottom triangular matrix with diagonal elements;
C is top triangular matrix, is solved by the top relaxation -0.80
0.00 0.02 0.04 0.06 0.08
method. t, s
Fig.9. The analogue of fig.4 for another parameters of electric
x n 1 px n
(21) B D x n 1
D Cx b
n sсheme
1 p The results of computer simulation
The results of computer simulation of transient processes
where D is diagonal matrix; p is relaxation parameter. If 0 <
of single-phase inverter in the regime of sudden switching-
p < 1 than we receive the top relaxation method; if p = 0 than in voltage source u(t) = 110 V are presented. The regime is
we receive Seidel’s method. calculated for the following parameters: C = 0.02 F; Lwr = 0.5
H. The time function of primary (1) and secondary (2) vo-
ltages are shown in fig.3 and currents shown in fig.4. The
time function of inductance coil (1) and resistor (2) voltages
are shown in fig.5. The spatial distribution in fixed time of
magnetic induction and vector potential of electromagnetic
field on cross-section of 2-windings iron-clad transformer
body quarter is shown in fig.6-7. As an example there is
computated the device regime for another parameters: C =
0.001F; Lwr = 0.01 H. The results (analogues of fig.3-4) are
presented on fig.8-9
Conclusion
Fig.6. The spatial distribution of magnetic induction in fixed time of 1. There is constructed mathematical model of
transient process (fig.1) calculation and research of electromagnetic processes of a
single-phase frequency divider on the basis of computation
of an electromagnetic field.
2. The constructed mathematical model allows, if
necessary, to realize calculation of operating duties of a
single-phase frequency divider and at the robot with inexact
corners of discovery of thyristors (in this case enough to
exchange the scheme of the description of operation of
semiconducting valves). Nevertheless practical
implementation of calculation in such case requires
considerable computing costs and HIGH-POWER
computer. At the same time the obtained information on
Fig.7. The spatial distribution of vector potential in fixed time of distribution of an electromagnetic field enables to realize
transient process (fig.1) calculation of dynamic parameters of the device in view of
200.0 commutation of valves.
3. The information on character of electromagnetic
processes, which flow past in the researched device,
100.0 enables to develop mathematical algorithm of calculation of
design data of devices of different power and for different
values of power with usage of optimization.
u1 , u2 , V
0.0 4. On the basis of the offered technique there can be
constructed mathematical models of other devices, in which
basic elements are the pole cores and windings.
-100.0
REFERENCES
[1] Tchaban V., Chromiak J., Peleshko D. Principle of field
-200.0 mathematical model construction of electrotechnical devices.
0.00 0.02 0.04 0.06 0.08 ISTS UEES’95, Szczecin, p.363-366.
t, s
[2] Tchaban V. Methods of non-linear electrotechnics (in Ukrainian).
Lviv, Svit, 1990, 190p
Fig.8. The analogue of fig.3 for another parameters of electric
sсheme Autorzy: prof. dr hab. Inz. Vasyl Tchaban, Rzeszow University,
Institute of Technology, Pilsudskiego 21/4 35074 Rzeszow, Lviv
polytechnic national university. E-mail: vtchaban@polynet. [Link];
vtchaban@[Link];
Dr hab. [Link] Peleshko of Lviv polytechnic national
university, Institute of Computer Sciences and Infromation
Technologies, E-mail: peleshko@[Link] .
PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 88 NR 6/2012 253