International Journal of Control and Automation

Vol. 9, No. 10 (2016), pp.361-372

http//dx.doi.org/10.14257/ijca.2016.9.10.33

Starting Strategy of Thyristor Power Supply in Parallel Resonant

Induction Heating on Heavy Load

Sen-Lin Cheng, Hu Xu, Bin Li, Bu-Qing Liu and Chuan Wang
School of Automation, Chongqing University, Chongqing, China
[email protected], [email protected], [email protected],
[email protected], [email protected]

Abstract
On heavy load, the energy supplement in the starting process of Thyristor Power
Supply for Parallel Resonant Induction Heating (TPSPRIH) has trouble in catching up its
consumption which makes the system start more difficultly. Aimed at the puzzle, this paper
proposes a starting strategy based on the analysis of parameters of heating inductor and
starting circuit. Considered the process complexity in commonly used pre-charged circuit,
it analyzed relationship among the parameters of heating inductor, the iron material and
its filling status. Compared the supply speed with the decay characteristic of load circuit,
the critical starting conditions of TPSPRIH on heavy load has been gotten, then derived
the solving space of trigger angle of rectifier on energy providing. Based on the rated
parameters of a typical system, the starting strategy of rectifier angle has been proposed.
The simulation on different regions in the solving space has been made. The simulation
result demonstrated the rationality and correctness. It probably provides a reference way
for the reliable starting of TPSPRIH.

Keywords: Induction Heating, Thyristor Power Supply, Starting Strategy, Heavy Load

1. Introduction
Induction heating power supply is an energy conversion device, which is widely
used in the forging industry [1, 2]. The TPSPRIH raised more concern because it
can provide more efficient power, but it’s difficult to start on heavy load because of
its big problem in its popularization and application, especially when it’s starting on
heavy load, the success rate of starting is extremely low [3-5]. Although many
scholars have studied on it, such as the [6] studied the pre-charge starting, the [7]
proposed the double auxiliary bridge starting, the [8] presented the load pre-
magnetizing starting etc., but their starting methods are mainly by means of
changing topology of circuit to increase the medium-frequency voltage when
converting the current so as to improve the success rate of starting on heavy load.
In fact, if the energy supply speed of load circuit is not high enough, the success
rate of starting is still very low although the supply voltage is at a high level. Study
in engineering found that, in the process of starting, the decay and supplying speed
of energy in load circuit, the state of load circuit [9-11] and the equivalent
parameters of heating inductor [12-15] in different starting conditions affect each
other, but a clearly comprehensive analysis of their interactio n has not been found.
In the paper, taken the commonly used pre-charge starting as an example, from the
view of rectifier control angle, it established the objective function, studied on the
characteristics of load circuit within the objective function on heavy load, and
exported the control conditions in the starting process of TPSPRIH.

ISSN: 2005-4297 IJCA

Copyright © 2016 SERSC
International Journal of Control and Automation
Vol. 9, No. 10 (2016)

2. Zero-Input Response on Different Load

The pre-charge starting is one of the forced starting methods, and there is a higher
success rate of starting on heavy load under the same condition [16]. The scheme circuit
of pre-charge starting is shown in Figure 1. T1 ~ T5 are thyristors, Ld is the inductance of
reactor in DC side. L, R, C, Id, iH and uH is respectively the equivalent inductance of
heating inductor, the equivalent resistance of heating inductor, the compensation capacitor
of load circuit, the current in DC side, the medium-frequency current and voltage of load
circuit.
During starting process, the pre-charged CP through T5 charges to the compensation
capacitor C. Before the initial energy of load circuit decay to zero, the triggering thyristor
T1 ~ T4 conducts alternately, and at the same time, it triggers the rectifier bridge. When the
medium - frequency voltage is established steadily on load circuit, it means to be started
successfully.

Rectifier iLd Ld
U Bridge
T1 T2
L iL R
iH C
V Ud DC Side
uH
CP T5
W Pre-charge
T4  Circuit + T3

Figure 1. The Scheme Circuit of Pre-Charged Starting For TPSPRIH

In the above equivalent load circuit, the L and R of equivalent heating inductor are
varied with the time varying. When it’s starting on heavy load, the varying rate of R and L
owns rather difference, it accelerates the energy decay speed on load circuit, it leads the
energy decay speed to be far higher than its supplying, and therefore, it results in thyristor
inverter becoming hard to convert current normally. The medium-frequency voltage on
the load becomes hard to establish, and the start will be failed.

500
uH(t)/V
400
Curve-1
300
200
100
0
-100
-200
-300
Curve-2
-400
t/s
-500
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

Figure 2. The Decay Curves on Different Load

The comparison response curves of load voltage in zero-input are shown in Figure 2. In
the figure, the curve-1 is the decay curve of load voltage on light load, and the curve-2 is
on heavy load. Both of them are at the same initial condition, and in which, their pre-
charged voltages are 500V, but the decay speed of curve-2 is faster than curve-1

362 Copyright © 2016 SERSC

 International Journal of Control and Automation
Vol. 9, No. 10 (2016)

obviously. The basic reason is that the variation rate of R and L is enormously different on
heavy load, which causes the decaying coefficient of load circuit to be increased greatly,
and therefore, it becomes more difficult to start.

3. Selection of Control Parameter

3.1. Performance of Heating Inductor
Heating inductor can be simplified as coil with iron core, and the equivalent circuit is
shown as in Figure 3. Where, RCu, RFe and L is respectively the wire resistance, the
equivalent resistance and equivalent inductance of heating inductor. Generally, the RCu is
mall enough to be ignored, and therefore, RFe and L are the key factors affecting the
technical performance of heating inductor.

iL iL RCu L iL L

u u RFe u R

Figure 3. Equivalent Circuit of Heating Inductor

When the iron bar is heated, it is put into the heating inductor. The equivalent
inductance and resistance is respectively shown as formula (1) and (2).
L  N r 0 n r 2 (1)
R  RFe (2)
Among them, n  N / l , l, r, μ0 and μr is respectively the turn number of coil in unit
length, coil length, bar radius, permeability of vacuum and the differential permeability of
iron bar. The μr - T curve is shown as in Figure 4, and in which, T0 is the Curie
temperature of iron bar.
μr
μrmax

μr=1
0 T0 T
Figure 4. μr – T Curve

From Figure 4, when the heating inductor is unloaded, there must be μr = 1. When it’s
on heavy load, it exists μr = μrmax. When it’s on heavy load with high temperature (i.e. T
≥ T0), it must meet μr = 1. From formula (1), when it’s on heavy load to start with the
physical size and other parameters of the heating inductor under the same condition, its
equivalent inductance L reaches to the maximum, and also the variation law of L-T is the
same as μ - T shown as in Figure 4.
Assuming that the turn number of heating inductor is N, and its current is iL(t) shown as
in Figure 3, then the magnetic induction on the coil axis is expressed as formula (3).
B   NiL (t ) (3)

Copyright © 2016 SERSC 363

International Journal of Control and Automation
Vol. 9, No. 10 (2016)

Within the heating inductor, before the temperature of heating material reaches its
Curie temperature T0, the iron loss includes the hysteresis loss and eddy current loss [17],
and the power of hysteresis loss part Ph can be expressed as following.
Ph  Kh  BmnV (4)
The eddy current loss part Pe is expressed as formula (5).
Pe (t )  Ke  2 Bm2V (5)
Where, ω, Bm, n, V, Kh, Ke is respectively the alternating frequency of magnetic field
which is the same as the load oscillation frequency in the paper, the amplitude of
magnetic induction intensity which is varying with formula (3), the exponential, the
volume of heating material, the correlation coefficient of ferromagnetic materials which is
determined by electrical resistivity, cross sectional area and shape which can be gotten
from experiment.
Combined formula (4) and (5), the iron loss can be expressed as formula (6).
PFe (t )  K Fe  Bm2V (6)
Where, KFe and β are the constant, in which, KFe is determined by experiment, and
generally, the β takes the value from 1.2 to 1.6.
When the temperature of heated material in the heating inductor is higher than its Curie
point T0, the iron loss only includes the eddy current loss shown as formula (5). In the
formula (6), when the iron loss and iL(t) are instantaneous value, it can get the resistance R
of heating inductor in DC circuit according to the thermal effect of resistance.
K   B 2V
R  RFe  Fe 2 m (7)
iL (t )
From formula (3), (6) and (7), it can be obtained that when the varying law of iL(t),
heating material and initial trigger frequency of inverter are the same, the equivalent
resistance R of heating inductor owns the relationship R∝Nμ2, and the equivalent L owns
the relationship L∝μ from formula (1). Now that the variation rate of R is much faster
than L under the same limit condition, the decay factor of load circuit will increase rapidly
because of the variation of heating inductor parameters R and L, and therefore, it will lead
to the fact that load energy decay is faster, and the starting on heavy load becomes more
difficult.
But the formula (7) is only a kind of qualitative estimation of the variation of R, and in
which, it is not applicable in engineering. Therefore, it is relatively accurate to the
estimating value of R based on the identification of oscillation frequency ω in the
beginning of starting process.
1
R  2L  2 (8)
LC

3.2. Analysis of Load Circuit

3.2.1. Current of Load Circuit

In the process of pre-charge starting, the initial trigger frequency of inverter is
determined by the natural oscillation of load circuit. When it’s on parallel resonance, the
load current in Figure 1 must meet the expression (9).
iL (t )  QiLd (t )sin t (9)
Where, Q is the quality factor of RLC parallel resonant circuit, and it can be expressed
as formula (10) approximately.
L
Q (10)
R
In the starting process, iL(t) consists of pre-charge decay oscillation current and
supplying current of DC side, i.e.

364 Copyright © 2016 SERSC

 International Journal of Control and Automation
Vol. 9, No. 10 (2016)

iL (t )  iL0 (t )  QiLd (t  t0 )sin (t  t0 ) (11)

Where, iL0(t) is the pre-charge decay oscillation current, namely the current of the pre-
charged energy to decay oscillation. iLd(t) is the supplying current of DC side, and t0 is the
lag trigger time of inverter.
It’s necessary to analyze the decaying process of pre-charged energy when the load
circuit is on zero-input state. And the load equivalent circuit is shown as the Figure 5. The
load circuit energy consumption is totally represented by the power of resistance R, so the
energy consumption speed of load circuit is the power on R.
S iL

R
C uH

Figure 5. Equivalent Load Circuit of Zero-Input Response

In Figure 5, the initial pre-charged voltage of compensation capacitor C is UC0, after

closing switch S, the parameters of the circuit must meet R  2 L / C , and the closed loop
of RLC would produce the decay oscillation.
d 2uH (t ) du (t )
LC 2
 RC H  uH (t )  0 (12)
dt dt
duH (t )
iL 0 (t )  C (13)
dt
From (12) and (13), the formula (14) would be held.
U C 0  t
iL 0 (t )  e sin t (14)
L
In which,
1
  2 (15)
LC
R
 (16)
2L
Where, UC0, ω and δ is respectively the pre-charged voltage of C, frequency of medium
- frequency decay oscillation, and the decay factor of medium - frequency voltage
decaying which reflects the consumption speed of energy of load circuit.
In the starting process, assuming that the rectifier triggering time is the initial moment
(t = 0), then the current iLd(t) through the reactor Ld can be expressed as formula (17).
t /
iLd (t )  I d 0  I F (1  e ) Ld
(17)
In which,
U
IF  d (18)
R1
L
 Ld  d (19)
R1
U d  U 2 f ( ) (20)
L
R1  (21)
RC
Where, Id0, IF, U2, Ud, α, f (α) and R1 is respectively the initial current of Ld, expected
steady state current, input phase effective voltage of rectifier, average voltage of rectifier
output, trigger angle of rectifier, function reflecting the relationship of Ud, rectifier angle,
and the equivalent oscillation resistance of load circuit.

Copyright © 2016 SERSC 365

International Journal of Control and Automation
Vol. 9, No. 10 (2016)

Integrating the formula (8), (11), (14) and (17), the current on heating inductor in
starting process can be expressed as formula (22).
U U f ( )  ( t t )/
iL (t )  C 0 e t sin t  Q[ I d 0  2 (1  e )]sin  (t  t0 )
0 Ld
(22)
L R1

3.2.2. Power of Load Circuit

In the starting process, the instantaneous power of load circuit is expressed as formula
(23).
Pout (t )  iL2 (t ) R (23)
Combining with formula (22), it would get the expression (24).
U U f ( )  ( t t )/
Pout (t )  ( C 0 e t sin t )2 R  {Q[ I d 0  2 (1  e 0
)]sin  (t  t0 )}2 R (24)
Ld

L R1
Where, the front part of Pout(t) is the consumption power of pre-charged energy, the
latter one is the consumption of supplying energy, and compared with the latter, the front
is negligible. Setting the trigger time of rectifier as t = 0, expression (25) and (26) are the
simplification of formula (11) and (24).
iL (t )  QiLd (t )sin t (25)
 L
2 2
U f ( )  t /
Pout (t )  [Id 0  2 (1  e Ld
)]2 sin 2 t (26)
R R1
And the instantaneous supplying power of load circuit from DC side can be expressed
as formula (27).
Pin  U d iLd (t ) (27)
Combining formula (17) with (20), the formula (28) can be derived.
U f ( )  t /
Pin  U 2 f ( )[ I d 0  2 (1  e Ld
)] (28)
R1

3.3. Starting Control of TPSPRIH

There are many factors which could affect the starting process, but the initial current Id0
of the DC side, rectifier angle α and the initial trigger frequency ω1 of inverter are the
most concerned factors in this paper.

3.3.1. Selection of Initial Current Id0

Because the initial current is too small, it would lead to start failure of the TPSPRIH on
heavy load, and thus, it is necessary to select an appropriate initial current. There must be
two crossing points in half-oscillation-period as Pout(t) = Pin(t) in engineering, or there will
be no difficulty in starting. Assuming that they crossed the second time at t = t1. If the
required voltage in the inverter convert process is U0, then it can get the time t1
concerning the limit condition.
1
t1  TH (29)
4
2
T (30)

Where, TH is the load oscillation period. And the medium frequency voltage of load
circuit is expressed as formula (31).
di (t )
uH (t1 )  L L  RiL (t1 ) (31)
dt t t 1

The thyristor converter must satisfy the basic conditions as formula (32) when
converting current.

366 Copyright © 2016 SERSC

 International Journal of Control and Automation
Vol. 9, No. 10 (2016)

uH (t1 )  U 0 (32)
From formula (9), (21), (25) - (26) and (28) - (32), the minimum value of Id0 is
expressed as formula (33).
L

2 RLd C
U 0 RLd  U 2 f ( ) ( R Ld C  L e
2 2
)
I d 0 min  (33)
RLd L

3.3.2. Selection of Rectifier Angle α

When starting on heavy load, the energy supplying speed is crucial, at the same time,
which is determined by α. Usually it requires that, the smaller the angle is, the better the
start performance. But the step voltage applied to the whole system is too large when α is
small, it would produce instantaneous high voltage in the reactor which could be up to
thousands or even tens of thousands volt. And it would break down the circuit system, on
the other hand, if α is too large, it can not meet the requirements of starting. Therefore it is
necessary to select an appropriate α to ensure the starting under the higher performance.
During the starting process, compared with the supplying energy, the pre-charged
energy can be ignored. Thus the key to start successfully lies in the energy supplying
speed, and in which, it can not less than its consumption speed.
Pin  Pout (34)

ud ( )  6U 2 sin(   ) (35)
3
Where, ud(α) is the instantaneous voltage of three-phase rectifier output when its
trigger angle is α. Taking Pin = Pout, and when α is taken as the maximum value based on
formula (26) and (28), it could get the Eq. (36).
 2 L2
ud ( max )iLd (t )  [iLd (t )]2 sin 2 t (36)
R
The solution of formula (36) can be expressed as formula (37).
   2 L2 
   arc sin[ iLd (t )sin 2 t ], 0  
 3 6 RU 2 6
 max   (37)
 2  arc sin[  L i (t )sin 2 t ],  7
2 2
 
 Ld
 3 6 RU 2 6 6
From formula (37), it can be derived that the maximum rectifier angle is determined by
resonant frequency, current iLd(t) in DC side, and the parameters of load circuit. The
above condition can ensure that whether it would be started successfully, but the starting
performance is not necessary to be the optimal. Therefore, it is necessary to explore the
optimal solution of α in the starting process, and which is decided by the designed safety
voltage of the power system. If the safe voltage is Us0 , the most voltage must be
generated on the reactor in the changing process of current iLd(t).
Assuming Id0 = 0 in the formula (17), and combined with formula (19), it could derive
out the change rate of current.
t
diLd (t ) ud ( )  Ld
 e (38)
dt Ld
And at the same time, the instantaneous voltage of reactor should be uL(t).
di (t )
uL (t )  Ld Ld (39)
dt
The limit of formula (39) is expressed as formula (40).

lim uL (t )  6U 2 sin(   ) (40)
x 0 +
3
There must own the expression as formula (41) under the safety consideration.

Copyright © 2016 SERSC 367

International Journal of Control and Automation
Vol. 9, No. 10 (2016)

uL (t )  U s 0 (41)
The minimum value of initial rectifier angle can be expressed by formula (40) and (41).
  U s0 
  3  arc sin , 0 
 6U 6
 min   2
(42)
 2  arc sin U s 0 , 
 
7

 3 6U 2 6 6
Formula (36) and (42) have given the limit conditions of rectifier angle when starting.
If αmax < αmin, it would be failure to start. Others, to enhance the starting performance, i.e.
to shorten the setting time, it requires the current iLd(t) to response at the most variation
speed all the time, in which, it gives out the maximum variation of α, namely Δαmax
expressed as formula (43).
 
  U s 0  6U 2 sin(   0 )
3 
    0  arc sin , 0   0   max 
 3 6U 2 6
 max   (43)
 
U s 0  6U 2 sin(   0 )
 2    arc sin 3 ,

  0   max 
7
 0
 3 6U 2 6 6
Where, α0 is the former trigger angle of rectifier, and Δαmax is the maximum variation
of α in the next rectifier period.

3.3.3. Selection of Inverter Trigger Frequency

The inverter trigger frequency ω1 is also an important control variable in the TPSPRIH
system, the state of the equivalent load circuit is different under different ω1. When ω1 >
ω, the load circuit is in capacitive state, when ω1 < ω, it’s in inductive state, and when ω1
= ω, it’s in resistive state that is called the resonant state.
From the formula (6), when the heated iron bar is in no difference in material and the
size, the load circuit power consumption is only determined by load oscillation frequency
because of Pout being positively correlated with ω. Therefore, if the frequency is selected
to be much smaller, the decay speed of energy will be largely lowered, and it owns more
advantageous to start. But concerned with the turn-off condition of thyristor, the
equivalent load circuit must be in capacitive state, thus ω1 > ω is a necessary condition
during the starting process. Under the comprehensive consideration, the reference
condition of inverter trigger frequency can be expressed as formula (44).
1  1.3 (44)

4. Simulation
4.1. The Range of Rectifier Angle
The selected region of rectifier angle is shown as in Figure 6. It is respectively selected
to the simulation value of equivalent resistance R and equivalent inductance L of load
inductor in terms of formula (8) and formula (1), the simulation value of Id0min and ω
refers to formula (33) and (44), the critical values αmax and αmin can be figured out in terms
of formula (37) and formula (42), and then the curves of αmax – δ and αmax – δ can be
plotted as in Figure 6.

368 Copyright © 2016 SERSC

 International Journal of Control and Automation
Vol. 9, No. 10 (2016)

80
/°
70

60
max
50

40 A
30
D min B
20
E
10 C

0
0 20 40 60 80 100 120 140 160

Figure 6. Variation of Curve Α- Δ

From Figure 6, it can be seen that the curves αmax – δ and αmax – δ divide the plane into
three parts. In which, the part C is safe starting region of rectifier angle. When the rectifier
angle is in the part D, it can not be started any longer. The part E is non-safe starting
region. A, B is exactly on the critical curve of αmax. When the load decay factor δ is
smaller, the range of α is wide. With the change of the load parameters, δ increases much
faster, and therefore, the range of α becomes an extremely narrow region.

4.2. Simulation Experiment of Starting Process

In the simulation experiment, the rated parameter of medium frequency power supply
is set as 1000KW/1000Hz, and it adapts MATLAB/Simulink as the simulation tool. The
simulation parameters are set as U2 = 254V, Ld = 5mH, C = 300µF, U0 = 20V, UH0 =
500V. The response curve of varying process of load voltage is shown as in Figure 7.

500 500
uH(t)/V uH(t)/V
400 400
300 300
200 200
100 100
0 0
-100 -100
-200 -200
-300 -300
-400 -400
t/s t/s
-500 -500
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

(a) Starting process of point A (b) Starting process of point B

Copyright © 2016 SERSC 369

International Journal of Control and Automation
Vol. 9, No. 10 (2016)

1,000 500
uH(t)/V uH(t)/V
800 400
600 300
400 200
200 100
0 0
-200 -100
-400 -200
-600 -300
-800 -400
t/s t/s
-1,000 -500
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

(c) Starting process in region C (d) Starting process in region D

Figure 7. Load Voltage Response in Starting Process

4.3. Analysis of Simulation Result

Figure 7 (a), (b) is respectively the load voltage response curve at point A and B in
critical rectifier angle of different maximum in Figure 6. Obviously, both situations could
start successfully with different α, but in the starting process, the instantaneous value of
load voltage is very close to the critical detecting value U0 which is set in simulation. At
the same time, the response speed gets slow, the regulation time of steady state gets
longer in time. In addition, the decay speed of A is smaller than B, but the starting process
is similar because both of them are under critical states.
Figure 7 (c) is the response of state for C in Figure 6. It’s in the starting region, also the
rectifier angle could not reach the critical value, and therefore, the starting performance is
superior to critical state.
Figure 7 (d) is the response of state for D in Figure 6. It’s in the cannot-start region.
Because the rectifier angle is lager than the critical value, the energy supply speed cannot
reach the requirement, and at last, it gets the failure to start.

5. Conclusions
In the pre-charge starting of TPSPRIH, the starting characteristic on heavy load is
determined by ① initial current of DC side, ② rectifier angle, and ③ initial trigger
frequency of inverter under other conditions being same. All of them would influence the
energy supplement, but the effect and start time are different. The upper limit of the initial
energy in the system depends on the condition ① , and it mainly determines the
circumstances before triggering. If the control of condition ② cannot be regulated in time,
it will affect the energy supply directly after the triggering, but a timely adjustment of
rectifier angle can improve the response speed according to the current state and the next
state of load. At the same time, the condition ③ will influence the commutation time of
inverter directly, which means a matchless trigger frequency leads to a failed
commutation of thyristor because of the complexity of starting process and the rapid
change of parameters, and thus, it is necessary to own a sufficient margin. The three
conditions mentioned above are expressed in explicit form, and from these conditions, it
can obtain unique solution at different state time.
At present, the most of TPSPRIH adopt the digital control type, and it will be more
convenient to express these constraints conditions with accurate value in engineering.
Therefore, it can obtain a better control performance, which makes a contribution to the
starting control about relevant AC–DC–AC power supply.

370 Copyright © 2016 SERSC

 International Journal of Control and Automation
Vol. 9, No. 10 (2016)

References
[1] I. Yilmaz, M. Ermiş and I. Çadirci, “Medium-Frequency Induction Melting Furnace as a Load on the
Power System”, IEEE Transactions on Industry Applications, vol.48, no.4, (2012), pp.1203-1214.
[2] A. Namadmalan, S. Abedi, S. H. Hosseinian and J. S. Moghani, “Power quality improvement for single
phase and three phase current source induction heating systems” Power Qual. Conf., PQC. 2010 1st
Power Quality Conferance Tehran, Iran, (2010).
[3] C. Saichol, K. Choyant and M. Veerapol, “Full-bridge current-fed inverter with automatic frequency
control for forging application”, IEEE TENCON 2004-2004 IEEE Region 10 Conference: Analog and
Digital Techniques in Electrical Engineering, Chiang Mai; Thailand, (2004), pp.D128-131.
[4] J. Jordán, J.M Magraner, C. Cases, V. Esteve, E. Dede, E. Sanchis, E. Maset, A. Ferreres, J.B. Ejea and
E. Gumbau, “Short-Circuit Critical Frequency for Induction Heating Parallel Resonant Inverters”, 13th
European Conference on Power Electronics and Applications, Barcelona, Spain, (2009).
[5] R. Fuentes, J. Juliet, F. Campaña, F. Ahumada, C. Silva and J. Estrada, “Comparative Analysis of Three
Starting Methods for Parallel Resonant Current Source Inverter”, 5th IET International Conference on
Power Electronics, Machines and Drives, Brighton, United Kingdom, (2010).
[6] I. Wallace, A. Bendre, J.P. Nord and G. Venkataramanan, “A unity-power-factor three-phase PWM
SCR rectifier for high-power applications in the metal industry”, IEEE Transactions on Industry
Applications, vol.38, no.4, (2002), pp.898-908.
[7] Q. Fu and M. Luo, “Technique of Starting Parallel Inverter Intermediate-Frequency Power Supply on
Heavy Load”, Transaction of Beijing Institute of Technology, vol.28, no.3, (2008), pp. 241-244.
[8] Z. Q. Meng, Y. Z. He and Y. Y. Peng, “A Study of the Starting Method of Magnetizing to the Load of
SCR Medium-frequency Power Supply”, Transactions of China Electrotechnical Society, vol.18, no.2,
(2003), pp. 57-61.
[9] M. G. Kim, G. Y. Park, D. H. Yoo and G. Y. Jeong, “High Power Full-Bridge DC-DC Converter using
a Center-Tapped Transformer and a Full-Wave Type Rectifier”, International Journal of Control and
Automation, vol.7, no.4, (2014), pp.267-278.
[10] Y. Sun, Y. H. Wang, X. G. Yang and L. L. Pang, “A Novel Coil Shape for Transverse Flux Induction
Heating”, Transactions of China Electrotechnical Society, vol.29, no.4, (2014), pp.85-90.
[11] J. S. Gao, F. X. Liu and Z.H. Sun, “Research on a Network Model of Induction Heating Loads”, Power
Electronics, vol.43, no. , (2009)1, pp. 39-41.
[12] J. G. Li, Y. R. Zhong and Z. F. Ming, “A New Analysis and Design method of Load-Matching Static
Characteristics for Voltage-Source Induction Heating”, Transactions of China Electrotechnical Society,
vol.21, no.11, (2006), pp.101-105.
[13] M. Xue, Q. X. Yang, Y. Li, X. Zhang and W. N. Liu, “The Study of Load Characteristic in Wireless
Energy Transfer System Based on EM Coupling Resonance”, Transactions of China Electrotechnical
Society, vol.28, no.2, (2013), pp.28-34.
[14] Z. J. Zhang, X. F. Jin, H. M. Li and Y. L. Peng, “Research on Induction Heating Voltage Source Inverter
Load Matching”, Journal of North China Electric Power University, vol.37, no.5, (2010), pp. 52-56.
[15] J. W. Yue, H. M. Chen and Z. S. Wang, “Research on Ultra High Frequency Induction Heating Power
Supply Inverter Load Topology”, Journal of Mechanical & Electrical Engineering, vol.27, no. 6, (2010),
pp. 73-77.
[16] G. Q. Fu, Y. H. Geng and E. H. Liu, “Two Kind of Startup Mode Design and Analysis of the
Intermediate Frequency Power Supply”, Journal of Hebei University of Technology, vol.36, no.4,
(2007), pp.48-51.
[17] Y. B. Han, “The Design of the Inductor Used in High-Frequency Induction Heating Equipment”, M.S.
of Zhengzhou University, (2013).

Copyright © 2016 SERSC 371

International Journal of Control and Automation
Vol. 9, No. 10 (2016)

Authors

Sen-Lin Cheng, he received his B.S., M.S. and Ph.D. degrees in

Detection Technology and Instrument, Automatic Control Theory
and Application and Mechanical Manufacture and Automation from
Chongqing University, China, in 1990, 1993 and 1999, respectively.
He has been an Associate Professor and supervisor for master student
in School of Automation, Chongqing University. He has also been a
member of fuzzy information and engineering branch of ORSC and
NDT Society. His research interests include power electronics,
pattern recognition and image technology, detection technology and
automation systems, remote sensing and control technology, and
embedded systems as well as network control.

Hu Xu, he received his B.S. degree in Automation from Dalian

Maritime University, China, in 2013. He is currently working toward
the M.S. degree in Control Science and Engineering at Chongqing
University. His research interests include power electronics, circuits
and systems and control systems.

Bin Li, he received the B.S., M.S. and Ph.D. degrees in Industrial
Automation, Automatic Control Theory and Application and
Mechanical Manufacture and Automation from Chongqing
University, China, in 1985, 1992 and 1999, respectively. He has been
an Associate Professor and supervisor for master student in School of
Automation, Chongqing University. His research interests include
control theory, navigation, guidance and control.

Bu-Qing Liu, she received her B.S. degree in Detection and

Control Technology from Nanchang Hangkong University, China, in
2013. She is currently working toward the M.S. degree in Control
Science and Engineering at Chongqing University. Her research
interests include tracking telemetering and command, wireless sensor
networks and cognitive radio.

Chuan Wang, he received his B.S. degree in Automation from

Chongqing University, China, in 2014. He is currently working
toward the M.S. degree in control engineering at Chongqing
University. His research interests include power electronics, circuits
and systems.

372 Copyright © 2016 SERSC