2015 IEEE PES Innovative Smart Grid Technologies Latin America (ISGT LATAM)

Design of model-based controllers applied to a solid-

state low voltage dc breaker
Juan D. Valladolid
Universidad Politcnica Salesiana
Dept. of Automotive Engineering
Calle Vieja 12-30, Cuenca, Ecuador
Ismael Minchala Mary J. Vergara
Universidad de Cuenca Universidad de los Andes
Dept. Electrical and Electronic Engineering Dept. of Technology and Design
Av. 12 de Abril, Cuenca, Ecuador Ncleo la Hechicera, Mrida, Venezuela

AbstractThis paper presents the methodology of design of forced commutation of the silicon-controlled rectifier (SCRs)
model-based sliding mode control (SMC) algorithms applied to in a circuit breaker. Reference [6] present a dc breaker using
power electronic dc-dc converters, which are part of the superconductors in which special cameras diminish the arc
components of a solid-state low voltage dc breaker (SLVDB). The voltage caused by the disconnection of the breaker terminals.
power converters used in the tested schemes of the SLVDB are
the boost and sepic dc-dc converters. Accurate disconnection
times, user-configured, are achieved with the proposed 800
controllers, as well as a complete minimization of the transient Transient recovery voltage
recovery voltage (TRV) in the breaker terminals. Details of the 600
performance of two SLVDB configurations are analyzed and
400
compared in order to establish the best design comprising
complexity vs. performance. MATLAB simulations support the 200
results and provide a reasonable picture of the operation of the
SLVDB.
Voltage[V]

Index Terms- dc breaker, fault current, sliding mode control, -200

transient recovery voltage. -400

-600
I. INTRODUCTION
-800

The advent of cutting edge technology for integrating -1000

renewable energy sources (RES) into the grid, has led to an
open research area in the microgrids technology [1]. RES are
becoming increasingly more accessible at the distribution
system level, through de-regulation policies, and costs droop.
Intermittent operation of RES leads to variations in frequency
and voltage magnitude in a power system, which could cause
stability issues in the system, transient and harmonic problems.
Frequency and voltage regulation in microgrids with RES are
main control challenges in the operation of these systems [2].
Figure 1 shows the TRV that arises at the terminals of a
circuit breaker after clearing a fault current in a highly
inductive circuit. TRV is one of the biggest transient issues in
devices that perform circuits disconnections. The TRV is a
decisive parameter that limits the ability of interruption of a
circuit breaker, causing loss of insulation and eventually
forcing a restart of the circuit [4].
Reducing the negative effects of the TRV motivates the
search of new alternatives in traditional fields, such as:
protection schemes, power electronics, etc. [1][3]. Some
designs of dc breakers are presented in [5], [6], which
incorporate solid-state elements in their designs. Reference [5]
presents the design of intelligent controllers for controlling the
0.11 0.115 0.12 0.125 0.13 0.135 0.14 0.145 0.15 0.155 0.16
Time[s]
Fig. 1. TRV at the terminals of the circuit breaker
This paper presents the methodology of design of SMC
algorithms to control the triggering of the SCRs of two
different configurations of dc-dc power converts, which are
part of the design of an SLVDB. The main goal of the control
strategy is to diminish the negative effects of the TRV, e.g.
high transitory voltages over the breaker terminals, and
possible reconnections of the circuit when a tripping command
is applied to the breaker. An opposite voltage to the main
source is applied to the circuit, in order to take the fault current
to a value near to zero, and safely disconnect the circuit.
This paper is organized as follows: Section II presents the
SMC design for the two different breaker configurations.
Section III presents a stability analysis of the designs using the
Lyapunov direct method. Section IV presents simulation
results of the designs. Section V presents the conclusions of the
research.

978-1-4673-6605-2/15/$31.00 2015 IEEE 337

 2015 IEEE PES Innovative Smart Grid Technologies Latin America (ISGT LATAM)

II. CONTROLLER DESIGN Stability; and

An indirect control approach is used for the design of the Performance analysis.
SMC to be used in the power converters. An electrical current References [8], [9] describe the basis of SMC design, which
signal is adjusted for tracking a voltage reference signal (Vref ), have been considered for the development of our approach.
which defines the sliding surface S(x) and the control action u, A. SLDVB design using SMC for controlling a bosst power
as it is shown in Fig. 2. The fault current sensing circuit converter
provides information to the SMC algorithm, which establishes
The model (1) describes the dynamics of the dc-dc boost
the switching conditions of the dc-dc converter that leads the
power converter, which forms part of the breaker design (Fig.
fault current to a value close to zero, in order to open the main
3). The following state equations are obtained:
switch and clear the fault.
x2 E x
x1 u 2
L L L (1)
x x V x
x2 1 2 p u 1
C CRb CRb C

where u represents the control signal, x1 is the current in L, x2

Fig. 2. General architecture of the control system
The methodology of design of the SMC for the SLVDB
presented in this paper, comprises two state-space models of
the power dc-dc converters:
Second-order model of the boost converter (Fig. 3)
Fourth-order model of the sepic converter (Fig. 4)
is the voltage in the capacitor C, V p represents the main
source and Rb a limiting resistor.
An indirect control action is proposed for controlling the
circuit breaker of Fig. 3, as it is depicted in Fig. 2. In order to
achieve this purpose, the following sliding surface is used for
designing the model-based SMC:
S ( x) x1 Z (2)

where Z is a general function for the sliding surface. Using the

expression (2) in the system (1), it is obtained:

E ZL (3)
ueq 1
x2 x2

The function S(x) is called the sliding surface function.

The sliding mode exists, if it the following is fulfilled:

dS (4)
Lg S ( x) 0
Fig. 3. Boost breaker model dx

where Lg is the directional derivative of the field g(x).

The system equations under these circumstances become:

d ( x1 Z ) d ( x1 Z ) (5)
Lg S ( x) g1 ( x) g 2 ( x)
dx1 dx2

x2 x1 (6)
Lg S ( x) (1 Z ) Z 0
L C

where g1(x) g1 ( x ) and g 2 ( x) represent the field, g(x),of each

Fig. 4. Sepic breaker model state in Eq. (1).
The equivalent control action (3) and Eq. (6) fulfill the
transversality conditions, which ensure that the vector fields of
For each proposed scheme we are to describe: the model reach the surface S(x) [9].
A dynamic mathematical model; The relationship between the states is shown in the
Conditions of existence of the sliding mode control; following expression:

338
 2015 IEEE PES Innovative Smart Grid Technologies Latin America (ISGT LATAM)

The output variable x4 is related to the sum of currents by

x x2Vp
2
2 (7) the following expression:
x1
ERb
x42 (V p E ) x4 V p E (12)
( x1 x2 )
The variable x1 determines the sliding surface to x2 . ERb

Eq. (12) shows the relationship between the output voltage

B. SLDVB design using SMC for controlling a sepic power x4 and the sliding current of the system. Using the Template
converter
III. STABILITY ANALISIS
The dynamic model of the sepic power converter (Fig. 4) is
determined as follows: The Lyapunov direct method [9] is used for the stability
analysis of the SMC designs. The selected candidate Lyapunov
E x3 x4 x x function, V (x) is the following:
x1 u 3 4
L1 L1 L1 L
1 L1
1 (13)
x x x V ( x) S ( x) 2 0
x2 u 3 4 4 (8) 2
L
2 L2 L2
x1 x x In the case of the boost power converter model, it is
x3 u 1 2 obtained:
C1 C
1 C1

x1 x x x V x
x4 u 1 2 2 p 4
C2 C
2 C 2 C 2 C R
2 b C 2 Rb
0 S ( x)(Cx2 CZx2 LZx1 )( x2 E LZ ) (14)

x1 and x2 are the currents of the x x (Cx2 CZx2 LZx1 )( x2 E LZ )

where the variables (1 Z ) 2 Z 1
L C CLx2
x x
inductors L1 and L2 , respectively; 3 and 4 are the capacitor (15)
voltage C1 and C2 , respectively. Since Eqs. (14)-(15) fulfill the condition of Eq. (13)
The sliding surface selected for controlling this circuit is the stability is guaranteed.
sum of the currents of the inductors L1 and L2 to generate an In the case of the sepic power converter model, it is
indirect control action [10]. The equation of the sliding surface obtained:
can be expressed as:
L ( x x4 E ) L1 x4 (16)
S ( x ) ( x1 x 2 ) Z (9) 0 S 2 3
L1 L2
By applying Eq. (9) in the system (8), the equivalent
control equation is obtained:
Stability is guaranteed in Eq. (16) when x4 > E
L ( x x4 E ) L1 x4 L1 L2 Z (10)
ueq 2 3
( L2 L1 )( x3 x4 ) IV. SIMULATIONS RESULTS
The following values are used for the simulation of the
From Eq. (4), we have:
R
circuit shown in Fig. 3: L = 2.6mH, C = 1mF, b = 0,5, E =
d [( x1 x2 ) Z ] d [( x1 x2 ) Z ] V
100V, p = 100V .
Lg S ( x) g1 ( x) g 2 ( x)
dx1 dx2 The simulation of the sepic circuit (Fig. 4) uses the
d [( x1 x2 ) Z ] d [( x1 x2 ) Z ]
g 3 ( x) g 4 ( x) following component values: L1 =3mH, L2 =2mH, C1 =1.4mF,
dx3 dx4 C2 = 1mF, Rb = 0.5 , E = 100V, Vp = 100V.
1 1 Z Z Z Z
( x3 x4 ) ( x1 x2 ) 0 For a ramp function that reaches the current failure in a time
L1 L2 L1 L2 C1 C2 of T= 25ms, it is considered:
(11) 0.5 (17)
x2 100 I0
Equations (10)-(11) fulfill the transversality conditions. T

339
 2015 IEEE PES Innovative Smart Grid Technologies Latin America (ISGT LATAM)

x2 100 20 I f t (18)
20 Boost Controller
Applying (18) in (7) we obtain the following sliding 18
Sepic Controller

surface: 16

14
(100 20 I f t ) 2 (100 20 I f t )100
x1d (19)

Fault Current [A]

12
50
10

Eq. (19) is used to generate the sliding surface to control the 8

current in the inductor L. This action has a direct effect on the 6

output voltage of the boost model. 4
In the case of the sepic model, Eq. (12) is used to generate
2
the sliding surface:
0

0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05

Time (s)
(100 20 I f t ) 2 (100 20 I f t )(V p E ) V p E
( x1 x2 )
ERb Fig. 6: Response of the system under a transitory decrease of the fault-current
(20)

Fig. 5 shows the excursion to zero of the fault current

20 Boost Controller
through an opposite voltage applied in coordination with the Sepic Controller
controllers signals, according to the sliding surface Eq. (18). 18

16

14
Fault Current [A]

20 Sliding surface 12

18 10
16 8
14 6
12 4
Current [A]

10 2

8 0

6 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05

Time (s)
4
Fig. 7: Response of the system under a transitory increase of the fault-current
2

0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 20 Boost Controller
Time[s]
Sepic Controller
18
Fig. 5: Response control algorithms for achieving current fault excursion to
16
zero in 25 milliseconds
14
Fault Current [A]

Figs. 6 and 7 show the performance of the two SLVDB 12

configurations. The boost breaker is disturbed in the state x2 , 10

while the sepic breaker is disturbed in the state x4 . Fig. 6 8

shows the response of the current excursion to zero of the two 6

breakers configurations, when a disturbance in the fault 4

current (decrease) with duration of 2.5 ms is present. Fig. 7 2
shows the dynamical response of the circuits current
0
excursion to zero, when a disturbance in the fault-current in
0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
the opposite direction (increase), as the one shown in Fig. 6, is Time (s)
present.
Fig. 9: Response of the system with a permanent disturbance (increase of the
fault-current)

340
 2015 IEEE PES Innovative Smart Grid Technologies Latin America (ISGT LATAM)
Figs. 8 and 9 show the behavior of the system
under a permanent disturbance.
The SLVDB designed with the sepic converter
offered a better performance than the one designed with the
boost converter. Table I compares the performance of the
control algorithms for each SLVDB model for achieving a
current excursion to zero in 25 milliseconds.
TABLE I
PERFORMANCE COMPARISON BETWEEN MODELS BOOST AND SEPIC
Parameter boost
model
Response time 25 ms
Stabilization time1 5 ms
Overshoot1 15%
Stabilization time2 6 ms
Overshoot2 17%
Stabilization time3 4 ms
Steady state error3 4 A.
1
under a transitory decrease of the fault-current
2
under a transitory increase of the fault-current
3
with a permanent disturbance
V. CONCLUSIONS
The methodology of design presented in this paper for
SMC applied to the triggering of SCRs, which controls dc
dc power converters to interrupt fault currents in circuit
breaker configurations, showed good results in effectively
diminishing the TRV effects at the breaker terminals when a
disconnection action is performed in a dc circuit with
inductive components.
It is possible to select, user configurable, the excursion
time of the fault current to zero, as well as the sliding
surface to lead the current to a safe range to disconnect the
faulty section of the circuit. The two breaker configuration
presented in the paper, using the boost and sepic power
converters achieved these features. Although, the circuit
breaker design using the sepic power converter offered a
better performance when perturbations on the fault current
(increases and decreases) appeared.
The SLVDB using the sepic power converter offers a
good performance, which allows the replacement of the
power source E (Fig. 4) for a precharged capacitor with a
peak voltage at least twice the value of E. Additional
circuitry and control techniques are required for an effective
charge of the capacitor with this configuration, which was
not part of the design presented in this paper.
REFERENCES
[1] Yinliang Xu, Wei Zhang, Hug G., Kar S., Zhicheng Li,
Comparative Control of Distributed Energy Storage Systems
in a Microgrid, Smart Grid, IEEE Transactions vol 6, n 1,
September 2014, pp. 238-248.
[2] Vargas-Martnez Adriana, Minchala-Avila Luis Ismael,
Zhang Youmin, Garza-Castaon Luis Eduardo, and Acosta-
Santana Patricia, Fault tolerant controller design for a master
generation unit in an isolated hybrid wind-diesel power
system, Int. J. Robust. Nonlinear Control, vol 25, n 5,
March 2015, pp.761-772, doi:10.1002/rnc.3277
[3] Michael R. Lavelle, P.E, Micro-Grid Applications Exploring
micro-grids for building energy applications Solving the
alternative energy management problem, Lavelle Energy
LLC,USA, 2010
[4] Ebrahim A. Badran, M. A. Abd-Allah, Abdelsalam H.
Hamza, and Tamer Elyan, A Proposed Transient Recovery
Voltage Mitigation Technique for Generator-Circuit-Breaker
sepic Fed Faults, Journal of Electrical Systems, Vol. 9, n 1,
model March 2013, pp.79-85.
25 ms [5] Avila, L.I.M., Castan, L.E.G.; Ortiz, E.R.C., An
3 ms Intelligent Control Approach for Designing a Low Voltage
5% DC Breaker, Andean Region International Conference
4 ms (ANDESCON), 2012 VI , vol., no., 7-9 Nov. 2012, pp.163-
6% 166.
2 ms [6] Morishita Y., Ishikawa T., Yamaguchi I., Okabe S., Ueta G.,
1 A. Yanabu S., Applications of DC Breakers and Concepts for
Superconducting Fault-Current Limiter for a DC Distribution
Network, Applied Superconductivity, IEEE Transactions on,
vol.19, no.4, August 2009, pp.3658-3664.
[7] L. Martinez-Salamero, A. Cid-Pastor, A. El Aroudi, R. Giral,
J. Calvente and G. Ruiz-Magaz, Sliding-Mode Control of
DC-DC Switching Converters, IFAC World Congress, vol
18,n., September 2011 pp.1910-1913.
[8] Jean J. E. Slotine, Weiping Li, Fundamentals of Lyapunov
Theory, in Applied Nonlinear Control, New Jersey: Prentice
Hall, 1991, pp. 41-99.
[9] Sira-Ramrez, H., Marquez, R., Rivas-Echeverra, F., &
Llanes-Santiago O., Introduccin al control por regmenes
deslizantes, in Control de Sistemas no Lineales, Madrid:
Prentice Hall, 2015, pp. 307-330.
[10] Shi-bing Wang, Yufei Zhou, Iu H.H.C., Jun-ning Chen,
Complex Phenomena in SEPIC Converter Based on Sliding
Mode Control, Circuits and Systems, 2007. ISCAS 2007.
IEEE International Symposium on, vol., no., 27-30 May
2007, pp.2407-2410.
341