An Optimal Selection of Induction heater

Capacitance by Genetic Algorithm

Considering Dissipation Loss Caused by ESR

E. Daryabeigi *, C. Coauthor** and F. Thvili Pour Fard *, C. Coauthor *** and G.Arab *
*Graduate student of Electrical engineering department, Islamic Azad university, Najafabad branch/as well as;
Young Researcher Club, Isfahan, Iran
**Graduate student of Electrical engineering department, Islamic Azad University, Najafabad branch, Isfahan,
Iran
***faculty of engineering department, Shahrekord University, shahrekord, Iran
daryabeigi_e_e_e@yahoo.com , *, tahvili_po_e@yahoo.com, **, arab_gh@eng.sku.ac.ir ***

Abstract__ In design of a parallel resonant Induction
heating system, choosing a proper capacitance for the
resonant circuit is quite important. The capacitance affects
the resonant frequency, output power, Q-factor, heating
efficiency and power factor. In this paper, the role of
equivalent series resistance (ESR) in the choice of
capacitance is significantly recognized. Optimal value of the
induction heating resonance capacitor is achieved by used
genetic algorithm under voltage constraint for maximizing
the output power of an induction heater, while minimizing
the power loss of the capacitor and inverter switching
frequency at the same time. Based on the equivalent circuit
model of an induction heating system, the output power, and
the capacitor losses are calculated. The effectiveness of the
proposed method is verified by computer simulations.
Keywords__ induction heating, capacitance, genetic
algorithm, Switching frequency, equivalent series
resistance.
I. INTRODUCTION
Induction heating is widely used in metal industry
for melting or heating of thin slab in a continuous casting
plant because of good heating efficiency, high production
rate, and clean working environments. A typical parallel
resonant inverter circuit for induction heater is shown in
Fig. 1[11].
In the parallel resonant inverter, if the switching
frequency is close to the resonant frequency, higher
voltage is generated at capacitor bank [2]. However, due
to the limit in the voltage tolerance of the capacitor bank,
the inverter output voltage VC needs to be limited below
the rated voltage Vmax . One method of limiting VC is to
reduce the DC-link current I dc by increasing the firing
angle of the rectifier. However, it results in the decrease
of I dc , thereby the decrease of the output power. At the
mini-mill in a POSCO steel plant, eight induction heaters
of load commutated type were installed to heat thin slabs
for next milling process in the continuous casting plant.
The rated output power of each induction heater is 1.5
MW. However, there were two problems: One was the
insufficiency in the output power, and the other was the
frequent damages of the capacitor bank. Insufficiency in
output power was caused by a poor power factor of the
inverter. On the other hand, the damage to the capacitor
bank was due to a little voltage margin between VC
and Vmax , and it resulted in a large power dissipation in
the capacitor causing a high temperature rise. The
capacitance of the capacitor bank affects the overall
operating factors of induction heater such as resonant
frequency, Q-factor, efficiency, and power factor [8].
Hence, in this work, we propose a method of choosing an
optimal capacitance value C opt , which maximizes the
output power, and at the same time, minimizes the
capacitor loss [11].
An optimal value of the capacitance under the voltage
constraint is found with the help of Lagrange multiplier
[9], but this time in problem solution research is centered
to used the calculate advanced methods , such as neural
network , genetic algorithm etc.
An optimal value of the capacitance is found by
defining an object function that includes output power,
power loss, switching frequency and efficiency. At the
first approach an equivalent model of the induction heater
is developed based on previous works [1], [4], [8]. The
heating coil and slab is modeled as an inductance plus a
series resistance, and the capacitor bank is modeled as a
pure capacitance with an equivalent series resistance
(ESR) [10]. Then optimal capacitance is achieved by
select of proper object function [11].
II. SYSTEM ANALYSIS
A. Equivalent Circuit
In general, the heating coil and the load are modeled
as a transformer with a single turn secondary winding as
shown in Fig. 3(a). Almost all magnetic flux generated by
the induction coil (primary winding) penetrates into the
slab (secondary winding). Hence, in the secondary circuit
no leakage inductance appears and
 Fig.1. Block diagram of induction heating system [11]

the coupling coefficient is equal to one. The secondary

circuit can be moved to the primary part as shown in Fig.
3(b). The slab resistance RL for one turn coil is given by:

L 2(ω + 2b)
RL = ρ =ρ , (1)
A lδ
Where L and A are length and area of eddy current, l is
the effective length of the slab occupied by one turn coil
and b and w are defined in Fig. 2 [11], δ and ρ are skin
depth almost distributed over the surface of slab and
electrical resistivity of the material. Simplified equivalent
model for a transformer can be represented in Fig. 3(c) by
an equivalent inductance L eq and resistance R eq [1].
These equivalent parameters depend on several variables
including the shape of the heating coil, the spacing
between the coil and slab, the electrical conductivity and
magnetic permeability of the slab, and the angular
frequency of the varying current ω S .
Leq = L1 − A L2 , 2
Req = R1 + A RL ,
2
Where R1 denotes the resistance of the heating coil, R L
denotes the resistance of the heated slab, and
A = ωs LM / ωs2 L22 + RL2 .It is noted that
inductance of heating coil L1 is not affected by the
existence of the slab in the heating coil, since at about
1100C" temperature the permeability of the iron slab is
equal to that of air, i.e.,
μ = 4π × 10−7 (H/m) [11].
To represent the power dissipation in the capacitor
bank, it is modeled by a pure capacitance C and an
equivalent series resistance (ESR) R ESR It is noted that
R ESR is inversely proportional to
Fig.2. Model of slab for one turn coil [11]
Capacitance, hence, it is modeled as R ESR = k / C ,
where k is a coefficient of ESR ranged from
1.2 × 10 −4 < k < 1.5 × 10 −4 Ω.F .
B. Power Equation
A useful variable to calculate the power is total
impedance seen from the impedance of equivalent circuit
in Fig. 3(c) [8]. The total impedance is given by:
( Z C + Z ESR ) ⋅ ( Z L + Z R )
Zt =
(2) Z L + Z R + Z C + Z ESR
(4)
(ω kR + ωs Leq ) + j (ωs2 Leq k − Req )
(3) = s eq ,
ωs (CReq + k ) + j (ωs2 LeqC − 1)
Where Z L = jω s Leq , Z C = 1 / jω s C , Z R = Req , and
Z ESR = RESR = k / C .
the
The rectifier and H-bridge inverter of the induction
heater are represented by a square waved current source
whose magnitude is equal to the DC-link current I dc .
Therefore, the current source expanded in a Fourier series
is described as follows:
the
 of the induction heater Pout and the capacitor loss Ploss
are given by:

2
⎛ I ⎞
Pout (C ) = ⎜ L ⎟ Z eq
⎝ 2⎠
2
1 Z C + Z ESR
= I s2 Z eq (9)
2 Z L + Z R + Z C + Z ESR
2

8I dc2 ωs k − j
Req ,
π 2
k
ωsC ( Req + ) + j (ωs2CLeq − 1)
C
2
⎛I ⎞
Ploss (C ) = ⎜ C ⎟ Z ESR
⎝ 2⎠
2
1 ZL + ZR
= I s2 Z ESR (10)
2 Z L + Z R + Z C + Z ESR
2

8I dc2 ωsC ( Req + jωs Leq ) k

,
(d) π 2
k
ωsC ( Req + ) + j (ωs2CLeq − 1) C
C
Fig.3. (a) Based circuit of the induction heater. (b) Equivalent circuit of
the induction heater on a transformer concept. (c,d) Simplified Where I L , I C denote the peak of iL , iC respectively [5]. It
equivalent circuit of the induction heater [11].
is noted that Pout and Ploss are function of capacitance
∞
4 I dc C, since all the parameters except capacitance are known
is (t ) = ∑ sin nωst n=1,3,5 … (5) values in (9) and (10).
n −1 nπ

III. OPTIMAL CAPACITANCE FOR MINIMIZING

The first harmonic amplitude is equaled as follows: DEFINED OBJECT FUNCTION BY THE CLASSIC
METHOD
I s = 4 I dc / π (6) In the load commutated inverter, the switching
frequency of the inverter must be higher than the resonant
The current through Req and RESR are represented by frequency of the L-C load to guarantee commutation of
the thyristors [3]. Hence, for more suitable value for the
iL and iC , respectively. The phasor expression of iL and inverter while working close to the resonant frequency,
we let ω a = 1.1ω 0 , and then the voltage constraint is
iC are described as follows:
given by
V Z +Z (7)
I = C = C ESR I , 4
L Z +Z Z +Z +Z +Z s VC = Z t ( jωs ) ⋅ I s = Z t ( J 1.1ω0 ) ⋅ I dc ≤ Vmax , (11)
L R L R C ESR π

VC ZL + ZR Where VC the peak of is vC , Vmax is rated voltage of the

IC = = Is ,
Z C + Z ESR Z L + Z R + Z C + Z ESR
Where VC and I s are phasors of vC and is and
VC = Z t ⋅ I s . In Fig. 3(c), the power consumption is
accomplished by equivalent resistor Req and ESR
RESR of the capacitor bank. Therefore, the output power
(8)
capacitor bank, and Z t is total impedance of capacitor
bank and heating parts. One can see that Z t ( j1.1ω0 ) is
also a function of capacitance, since ω 0 = 1 / LW CW .
The aim is to find an optimal capacitance value that
maximize the output power of the induction heater and
minimize the capacitor losses simultaneously under the
 voltage constraint (11). If the object function is selected
as Pout − Ploss , maximizing Pout − Ploss is equivalent to
maximizing Pout and at the same time minimizing Ploss .
Maximizing Pout − Ploss
4
Subject to Z t ( J 1.1ω0 ) ⋅ I dc ≤ Vmax ,
π % Randman =
It leads to apply Kuhn-Tucker theorem [9]. The cost
function is defined by
J (C ) = Pout (C ) − Ploss (C )
4 (12)
+ λ.(VMAX − Z t ( J 1.1ω 0 ⋅ I dc ),
π in high powers, therefore in this position must had upper
λ ≥ 0 is Lagrange coefficient .maximum point is defined
by :
∂J (13)
= 0,
∂C
π and error proportional to a few optimal characteristic
λ ⋅ (Vmax − Z t ( j1.1ω0 ) ⋅ I dc ) = 0. (14)
4
But, the resulting equation is quite long, and also
subject of frequency is not representation with another
subject synchronously [11].
IV. IMPLEMENTATION OF GENETIC ALGORITM
FOR OPTIMIZATION AND its RESULTS
Select of a suitable performance index is extremely
important for the design of induction heating. The present
work considers a performance index that can be written in
the general form as in (15) and the optimal induction
heating parameters shall be obtained by minimizing J ,
and it's propose is select of suitable capacitance, such as
increase of output power, decrease of loss and also
voltage and switching frequency constraints are satisfied.
Where desired of parameters: 1-Ferequence f S (that with
attention to industrial making constraints of high
frequency inverters can't increases to each value specific
for high powers.) 2- efficiency 3- output power 4-
difference output power with power loss 5- capacitance
value (with attention to this pattern that higher
capacitance is more expensive and larger than other,
therefore must is used less capacitance within possible
limit). Now, coefficients and weights must be imputed to
functions of mentioned parameters that show its
importance rate.
1
p
( out −1)
J= + 0.995 75e3 f s
0.0005
( Pout (c) − Ploss (c)) 2 (15)
+ Ceq + 100
1
+ 0.87
p
( out −1)
75e 3
1
,
Pout Randman
Z t ( j1.1ω0 ) ⋅ I s ≤ Vmox (16)
Pout (c) (17)
× 100
Pout (c) + Ploss (c)
Minimization of J that is equal maximization
Pout − Ploss ,output power and defining suitable and
minimum values of f S and Ceq , maximum and suitable
efficiency. With attention to the said subjects about
mentioned parameters, frequency value had to decrease
weight, that effect of minimizing was more than position
of low power. (But, it should be attend that high
frequencies will be culminated more suitable efficiency,
output power and capacitance values, but have limitation
of generation the high frequency too) of course latter we
will show that increasing frequency effect to gain of
parameters directly and then it will not be very effect to
results. The weights numerical values are achieved by try
models.
V. SIMULATION RESULTS
Simulation was performed with MATLAB software.
The parameters of POSCO induction heater were utilized
in this simulation are given in table (1).
With the use of GA, the optimal capacitance value is
found to be CW = 9.144[ μF ] by minimization (15)
and attention to (16).GA parameters are given in table
(2). The result of GA performance is show in fig.(7) that
move to minimum value of object function J (15).
Curve of output power Pout versus switching different
frequencies f S of mentioned system is indicate in fig.(5),
which by attention to that we see by increase of
frequency, decreases the procedure of increasing output
power, such that the end of curve, with increasing
frequency, there will not any change for output power,
subsequences optimal and suitable frequency for having
high power at acceptable frequency (by attention to
previous mentioned subjects) the suitable place become
place which after that increasing f S not very effected in
Pout . This place is show at fig.(5,6). In fig.(6) is show
derivative of curve of Pout versus f S that shows the
procedure of increasing output power of effective
increasing frequency. After characterized point on the
curve, output power will not have so change.
Just as is said this place is same optimal point. This
paper presents novel method for optimal select
capacitance and suitable switching frequency, with
considerations of tolerance of voltage and frequency,
ESR of capacitance, maximum output power and high
power factor.
 TABEL I
PARAMETERS OF INDUCTION HEATING

Lw 19.949[ μH ]

RL 0.0908[Ω ]
VMAX 1700[V ]

Is 1300[ A]
K 1.35 × 10 −4 [Ω. f ]

Figur5- altering frequency with output power

Figur4- vc voltage of capacitance C=9.144E-6

TABEL II.
PARAMETERS OF GA
Figur6- d( Pout)/d(fs)
Population size 18

Mutation rate 0.15

generations 100

Number of Variable 1

Bit of each Variable 24

best object 1.01615480747

CW = 9.144[μF ]
best solution
(0.23427^8=C)
Foigur7- converge object function in GA

VI. CONCLUSION
This paper suggests a new method in the choice of
capacitance and operating frequency with the
considerations on voltage tolerance and ESR of the
capacitor, maximum output power, and high power
factor, switching frequency. The optimal solution is
found with GENETIC ALGORITM and defining an
object function that includes output power, loss power,
efficiency and switching frequency. This optimal choice
is thought to contribute to increasing the life time of the
capacitor bank and generating a maximum output power.
Results of simulation are emphasized the improve
efficiency and decreasing switching frequency versus
Lagrange multiplier.
REFERENCES
[1] Y. Kwon, S. Yoo, and D. Hyun, “Half-bridge series resonant
inverter for induction heating applications with load- adaptive PFM
control strategy”,IEEE-APEC Conf. Rec., pp. 575-581, 1999.
[2] F.P. Dawson, and P. Jain, “A comparison of load commutated
inverter systems for induction heating and melting applications”, IEEE
Trans. On Power Electronics, Vol. 6, No. 3, pp. 430 -441, July, 1991.
[3] R. Bonert, and J.D. Lavers, “Simple starting scheme for a parallel
resonance inverter for induction heating”, IEEE Trans. on Power
Electronics, Vol. 9, No. 3, pp. 281 -287, May, 1994.
[4] T. Imai, K. Sakiyama, I. Hirota, and H. Omori, “A study of
impedance analysis for an induction heating device by applying a new
interpolation method”, IEEE Trans. on Magnetics, Vol. 33, No. 2, pp.
2143-2146, March, 1997.
[5] S. Dieckerhoff, M. J. Ruan, De Doncker, “Design of an IGBT-based
LCL-resonant inverter for high-frequency induction heating”, Industry
Applications Conference, Conference Record of the 1999 IEEE, vol.3,
pp. 2039 -2045, 1999.
[6] H. Kojima, T. Shimizu, M. Shioya, G. Kimura, “High frequency
power transmission in plural induction heating load circuits”, Industrial
ElectronicsSociety, 16th Annual Conference of IEEE, vol.2 pp. 990 -
995, 1990.
[7] Harold A. Wheeler, “Formulas for the Skin Effect”, Proceedings of
the I.R.E., pp. 412-424, September 1942.
[10] S. Zinn and S. L. Semiatin, Elements of Induction Heating. Ohio:
Electric Power Research Institute, Inc. 1988.
[8] E. J. Davies, Conduction and Iduction Heating. Londonn: Peter
Peregrinus Ltd. 1990.
[9] D.G. Luenberger, Linear and Nonlinaer rogramming. Stanford
University: Addison-Wesley Publishing Company. 1989.
[10] Dec.2002 issue of Practical Wireless in World Wide Web,
“http://www.valveandvintage.co.uk/pw/esr.htm”
[11] Jung-gi Lee, "An Optimal Selection of Induction Heater
Capacitance Considering Dissipation Loss Caused by ESR" IEEE 2004