Download CONFERENCE-Implementation-and-design-of-Fuzzy

Implementation and design of Fuzzy controller for high
performance DC-DC boost converters
A. Mansouri, F. Krim
Dept of Electronics, Faculty of Technology, University of SETIF1, Algeria.
Laboratory of Power Electronics and Industrial Control (LEPCI)
[email protected], [email protected]
Abstract—This paper discusses the implementation and
design of both linear PI and fuzzy controllers for DC-DC
boost converters. Design of PI controllers is based on
temporal response of closed-loop converters, while fuzzy
controllers design is based on heuristic knowledge of
boost converters. Linear controller implementation is quite
straightforward relying on mathematical models, while
fuzzy controller implementation employs one or more
artificial intelligences techniques. Comparison between
these boost controllers is made in design aspect.
Experimental results show that the proposed fuzzy
controller system is robust against input voltage and load
resistance changing and in respect of start-up transient.
Results indicate that fuzzy controller can achieve best
control performance concerning faster transient response,
steady-state response good stability and accuracy under
different operating conditions. Fuzzy controller is more
suitable to control boost converters.
Key-words—Boost
controllers
DC-DC
converter,
Fuzzy,
PI
I. INTRODUCTION
Fuzzy logic has rapidly become one of the most
successful of today technologies for developing
sophisticated control systems. Fuzzy control is considered
one of the most important applications in fuzzy logic
which is a technique developed to imitate human
behaviour. Fuzzy logic control is one of the intelligent
schemes that convert linguistic control strategy based on
expert knowledge, human intuition and heuristics into
automatic control actions offering a high-level
computation. Over past two decades, fuzzy controllers
have been successfully developed and used in various
applications in power electronic converters as power
quality is a serious particular issue in the developing
countries [1][5]. Fuzzy logic controllers have recently
been implemented as embedded controllers for robotics
[6] and have been applied to a broad range in Engineering
Sciences [7], [8]. Numerous motor drive problems have
been solved using fuzzy principles [9]–[11].
Conventional DC-DC buck, boost, fly-back and cùk
(buck-boost) regulators are power electronic systems that
convert a fixed DC voltage source into a minimum loss
variable voltage by a switching technique using BJT,
MOSFET, IGBT transistors or GTO thyristors [12][17].
MOSFET are extremely popular in low power, low
voltage and high frequency switching applications
(hundreds of kHz).including switching mode small power
supplies, solid state DC relays and brushless DC motors.
DC converters are used extensively in personal computers,
computer peripherals, adapters of consumer electronic
devices, and in telecommunication where efficiency,
volume and mass constitute important constraint in
aerospace, undersea, airborne systems, military
equipments. IGBT are extremely popular devices used in
medium and high power applications to control DC
motors in the converter circuits, in DCAC drives in
inverter circuits, in high power supply systems. They can
be used for traction motor control in electric automobiles,
trolley cars, forklift trucks and mine haulers.
Control technique for DC-DC converters must cope
with their wide input voltage and load variations to ensure
stability under any operating conditions while providing
fast transient response, smooth acceleration control and
high efficiency. Most used conventional controllers in
industrial applications are PI because of their simple
structure and good performance in wide range of
operating conditions. However, control parameters for
design of an optimal compensation circuit for closed loop
operation of converters, which ensure proper behavior in
any operating conditions, are difficult to obtain because
dynamic of DC-DC converters is non-linear, and practical
converter operation deviates from theoretical prediction
due to problems associated with parasitic resistances, stray
capacitances and leakage inductances of the components.
Various analysing techniques are adapted as standard
instruments for modelling and control of DC-DC
converters using state space averaging method [18], but
linearization limits their validity to small signal behaviour
in operating point before others techniques [19], [20] such
as LaplaceTransform and Bode plots can be applied.
In recent years, artificial intelligence techniques could
be applied such as Artificial Neural Network (ANN)
control based on learning to simulate human thinking [21],
[22], Fuzzy Logic (FL) control requires linguistic
variables to model inaccurate or incomplete knowledge
[23][26], Fuzzy Neural Network (FNN) control [27],
[28] and their mixture with genetic algorithms (GA) in
direct optimization approaches [29].
To test dynamic properties and robustness of
intelligent controller under any operating conditions, this
paper proposes an adaptive fuzzy inference system (FIS)
based controller for DC-DC boost converters. In FIS
design, controller is able to regulate the output voltage to a
desired value without steady-state oscillations despite
change in load resistances or reference voltages, provided
fuzzy controller has been designed accurately. For fuzzy
control based on artificial intelligences, design concept is
totally different from conventional PI controller
techniques using a plant model. Then, operation of fuzzy
controller does not rely on how accurate model is, but on
how effective linguistic rules are, giving better
performance in such cases. This makes them more
effective in applications where existing models are
nonlinear, ill-defined and not reliable enough. Fuzzy
control therefore simplifies design of optimal
compensation for DC-DC converters.
II.
PI BOOST CONVERTER CONTROL
pulse width modulation (PWM) control signal for DC
converters. Continuous mode modelling is naturally very
rigorous and well adapted to DC converters operating in
PWM.
Second order OLTF of buck-boost converter small
signal model is given by the following equations (3), (4)
for standard state space averaging techniques where D is
the nominal duty cycle and D0=1-D:
'
V
V0 .R.D0  0 Ls
Vˆ ( s )
D0
(3)
H ( s ) BO  0

2
2
ˆ
RLC .s  L.s  RD0
d ( s)
'
Vˆ ( s)
V0 .R.D0
(4)
G( s) BO  0

2
2
Vref ( s) RLC .s  L.s  RD0
Major characteristics of typical control systems,
which are often used as measures of performance to
evaluate systems, are stability, accuracy, and speed of
response. Feedback control systems could present defaults
so as insufficient steady-state accuracy, bad stability, too
high overshoot, and a long settling time. For design of
control systems, it is often necessary to introduce
controllers to meet a set of specifications witch define the To control CL system, it is necessary to choose
overall performance parameters of systems. We will coefficients values of KP and Ki using trial and error
implement PI and fuzzy controllers to compare their procedure to meet a set of specifications which define the
overall performance because design of control system is
performances.
Controlled system must respond to the required an attempt. Optimal values of PI controller are:
KP = 0.75
Ki = 200
performances by industrial context characterising its
transient and steady states: transient state is characterized III. FUZZY CONTROLLER FOR DC-DC CONVERTERS
by damping factor (minimal overshoot) and settling time
We are interested to replace PI conventional regulator
(shorter is within 5% criterion faster response will be).
Steady state response is characterised by its static by fuzzy logic controller. For any application, block
accuracy and stability. PI regulator parameters will be diagram of fuzzy controller scheme shown in fig.2 of DCimplemented to compensate instability or poor stability, DC converters is composed of five elements computed in
insufficient precision, sluggish and large overshoot. PI three steps: Fuzzifier (1), Decision Making based on fuzzy
controller which combines two control actions can be logic rules (2), Defuzzifier (3).
represented by the following control law:
Rules base
(1)
u (t )  K p .e(t )  K i  e(t ).dt
e(t) G1
G3
Where e(t) is the output voltage error :
e(t )  Vref t   V0 t 
ce(t)
Fuzzifier (1)
Decision
Making (2)
Defuzzifier (3)
(2)
d
G2
Proportional gain control KP provides control signal
proportional to error. It acts as an amplifier.
- If KP is large enough, control signal increases rapidly:
overshooting risk and oscillation output increase.
- If KP is small, control signal increases slowly: no
oscillating risks.
Integral action gain Ki reacts slowly to changing error and
insures progressive compensation of the reference. While
positive error (or negative) subsists, the action u(t)
increases (or decreases) until error annulations.
Fig. 1 illustrates PI controller boost block diagram to
regulate its output voltage V0 and current I0.
Vref
PI controller
up
d
e
PWM
+
KP
+
-
+
u
i
Ki / s
V0 dn
_
Boost
V0 ; I0
Controller
_
Fig .1: PI buck-boost
- controller block diagram.
_
PI Regulator stabilizes the load output voltage and
_
then must eliminate error between reference voltage Vref
8 V0. It provides the duty cycle
and measured output voltage
d which is compared with- a saw tooth voltage to generate
Data base
Fig.2: Fuzzy controller scheme.
Fuzzifier (1): error e(t) and change of error ce(t) are input
physical variables of the fuzzy controller that will be first
normalized to an universe of discourse with scaling
factors, G1 and G2 gains respectively, for a control action
defined by change in duty cycle d scaled by a
normalized gain G3 permitting adaptive signals
processing. These elements act globally over the control
increasing or reducing universe of discourse of control
quantities. Fuzzifier module converts these normalized
inputs to linguistic codes or sets described by membership
factors e, ce, .d. Fuzzifier requires definition of
linguistic sets and commonly used membership functions
triangular, trapezoidal or bell-shaped depending on the
individual application and experience of the process user
[30]. These one will be defined over normalized [-1, 1]
universe of discourse to affect membership degree to
fuzzy input sets (fig.3a), divided into five triangular
fuzzy-set levels, and a normalized [0, 1] universe of
discourse for the output, partitioned in seven triangular
fuzzy-set values (fig.3b) [31].
Decision making (2): Fuzzy controller output d is
deduced from the Fuzzifier inputs e and ce using
individual contributions of each rule from the rules base to
form a decision table for the fuzzy controller. Table 1
represents the inference matrix composed of 25 control
rules showing different situations usually obtained from
expert knowledge or heuristic (based on judgments of
experienced plant operator) and expressed as a set of IfThen rules to obtain a non linear relationship between
system states and control action which gives better control
action than conventional techniques. For fuzzy control, we
can use one of the most used inference mechanism: MAXMIN, MAX-PROD, and SOM-PROD.
Defuzzifier (3): Fuzzy logic controller output expressed in
linguistic code must be converted to physical or numerical
values and scaled to determine the well adapted action of
the controller from all contributions. Defuzzifier can
either choose one of the most used algorithms: maximum
criterion or centre of gravity methods (COG) [32]. COG
method is the most popular defuzzification technique and
is widely used in actual applications because the easier
maximum method is inefficient when the membership
function posses many maximums corresponding to
maximum abscise of the output value. The weighted
average of the membership function or the COG of the
area bounded by the membership function curve is
computed to be the crispest value of the fuzzy quantity.
Error Fuzzy Set
Degree of Membership mu(E)
NB
1
NS
ZE
PS
PB
0.8
0.6
0.4
0.2
0
-1
-0.8
-0.6
-0.4
-0.2
0
Error
0.2
0.4
0.6
0.8
1
Fig.3.a Inputs Membership functions for e and ce.
Degree of Membership mu(Delda d)
Changing Duty Cycle Fuzzy Set
PB
1
PM
SS
M
PB
MB
BB
the error e(k) and the change of error ∆e(k), illustrating
clearly the nonlinear characteristics of the proposed fuzzy
controller structure.
Fig .3.c: Fuzzy control surface.
TABLE 1 : FUZZY CONTROL RULES.
en en NB
NB
NS
ZE
PS
PB
PB
PB
PM
PS
M
NS
PB
PM
PS
M
BP
ZE
PM
PS
M
BP
BM
PS
PS
M
BP
BM
BB
PB
M
BP
BM
BB
BB
Five fuzzy levels or sets are chosen and defined by the
following library of fuzzy-set values for the error en and
change in error en:
PS: Positive Small; ZE: Zero Equal; NB: Negative Big;
PM: Positive Medium; NS: Negative Small; PB: Positive
Big; BP: Big Positive NM: Negative Medium; M:
Medium; BM: Big Medium; BB: Big Big.
The number of fuzzy levels is not fixed and depends on
the input resolution needed in an application. Larger the
number of fuzzy levels, higher is the input resolution.
Inputs are not quantized in the classical sense that each
input is assigned to exactly one level.
Triangular functions are used here to reduce complexity
in calculations [33]. Hence, the fuzzy representation of
quantized values and are the fuzzy sets and the degree to
which they belong to each fuzzy set.
Inference matrix in table 1 is established by a logic
taking into account the system physics. Hence a perfect
knowledge of its behaviour to regulate permit establishing
fuzzy rules different from conventional methods based on
mathematical models. Fuzzy control action can be
expressed as follow:
0.8
d = If e is NB and e is NB then d is PB
if e is NB and e is PS then d is PS
0.6
Or
Or
0.4
Fuzzy control model of boost DC-DC converter
0.2
0
0
0.1
0.2
0.3
0.4
0.5
Delta d
0.6
0.7
0.8
0.9
1
Fig. 3.b Output Membership functions for d.
Fig. 3c hereunder represents the normalized output of the
proposed fuzzy boost controller structure as a function of
…………………………………
.
Fig.4 scheme describes block diagram of fuzzy logic
control configuration for the DC-DC boost converter. The
boost converter is represented by a black box from which
only the terminals corresponding to the output voltage V0,
DC voltage source Vref and the control duty cycle inputs
are shown. The fuzzy controller used to adjust the
converter output voltage is the Mamdani implication
technique. Inputs of the fuzzy controller are error e and
change of error ce. The output of the fuzzy controller is
the normalized variation of the duty cycle d, computed
in the following three steps for fuzzy adjusting.
DC voltage
source Vref
+
-
PWM
V0
Stage
Integrator
1
s
Gain
G3
Fuzzy
controller
dk
+
BOOST
Converter
Gains
G1
ek
cek
A. Simulation of Fuzzy and PI controlled boost converter
- Vref
Error
+
Derivative
G2
L= 0.038mH RL=0.47Ω, C=220μH, Load R=50Ω,
Vin=30v and V0=50v, PWM switching frequency fs=12
KHz.
Regulated DC
Output V0
dk
Duty cycle
Results are given for the simulation of a Boost
converter acting in continuous mode. Converter
parameters are:
1) Fixed input voltage Vref and load R
We fix Vref=50V and load R=50Ω values, to observe
system performance parameters: settling time ts, overshoot
Dmax in transient state; stability, accuracy in steady state.
Fig.5 (a, b, c, d) shows the controlled fuzzy and PI boosts
temporal responses of output voltage V0; output current I0,
duty cycle d and error E.
ek
Fig .4 Fuzzy controller boost block diagram.
Inputs of the boost fuzzy controller are:
1. Error e(k) (4) which is the difference between sampled
reference voltage Vref(k) and converter sampled output
voltage V0 (k):
th
2. Change of voltage error Ce(k) (5) at the k
instant is defined as follow:
Ce(k) = e(k)  e(k-1)
40
Vout (V)
(4)
50
30
Fuzzy V0
PI V0
20
sampling
10
0
0
(5)
Output dk of the controller computed, using the
preferred COG method, is compared to a wave tooth
signal to generate PWM control signal. This signal is
generated to control switch in the boost converter
appropriately, to adjust the desired level output voltage V0.
The three variables e, e, d are normalized so as the
scaling factors G1 =80, G2=1/80 and G3 were computed in
such a way that normalized inputs e(k) and e(k) are well
adapted to the [-1,1] universe of discourse and a
normalised one [0, 1] for the output for any operating
points, expressed as follow in (6):
0.05
a)
0.1
0.15
Time [s]
0.2
0.3
Fuzzy and PI boost Output voltages Vout .
1
0.8
0.6
0.4
Fuzzy I0
PI I0
0.2
ek= e  G1 ; ek = e  G2 ; dk =dk  G3
0.25
Fuzzy & PI boost Currents IO for fix Vref & R
1.2
Iout (A)
e(k) = Vref (k)  V0 (k)
Fuzzy & PI boost Voltages VO for fix Vref & R
60
(6)
0
0
Output of the fuzzy boost converter is the duty cycle dk
at the kTh sampling time determined by adding the
previous duty cycle dk-1 to the calculated change in duty
cycle dn scaled by G3, defined by (7):
0.05
b)
0.1
0.15
Time [s]
0.2
0.25
0.3
Fuzzy and PI boost Output currents Iout.
PI & Fuzzy boost Errors
60
dk =dk-1 + dk  G3
50
Where dk is the inferred change of duty cycle by the
fuzzy controller kTh sampling time, and G3 is the gain
factor of the fuzzy controller. Adjusting G3 can change the
effective gain of the controller.
In classical control theory, (7) represents an
integrating effect of the fuzzy controller output which
increases system type and improves steady state error.
40
IV.
RESULTS AND DISCUSSION
E [V]
(7)
PI Error
Fuzzy Error
30
20
10
0
-10
0
0.05
c)
0.1
0.15
Time (s)
0.2
Fuzzy and PI boost Errors E.
0.25
0.3
Fuzzy & PI boost Duty Cycles d for fix Vref & R
0.5
PI boost Error E for changing R & fix Vref
0.6
0.4
Error E (V)
0.5
d
0.3
Fuzzy duty cycle d
PI duty cycle d
0.2
0.4
0.3
0.2
0.1
0.1
0
0
0
0.05
d)
0.1
0.15
Time [s]
0.2
0.25
0.3
-0.1
0
0.05
0.1
Fuzzy and PI boost duty cycles d.
c)
Fig.6 Fuzzy and PI controlled boost responses for fix Vref and R.
0.2
0.25
0.3
0.25
0.3
PI boost error E.
PI boost Duty Cycle d for changing R & fix Vref
0.5
0.4
0.3
d
For PI boost controller, fig.5 show that along
transient state, overshoot value reaches Dmax=14.8%,
system oscillations are quickly dumped and the steady
state is reached after ts=22.5ms for a small steady state
error EPI=0.5623. Increasing of duty cycle is proportional
to output voltage rising and vice versa.
0.15
Time [s]
0.2
0.1
Fig 5 show fuzzy boost controller responses with an
overshoot Dmax=3.6 %, reaching very quickly steady state
final values in a settling time ts =5ms for a very small
steady error EFUZZY= 0.0821<Epi and a duty cycle d= 0.42.
0
0
0.05
0.1
d)
2) Changing load R and fixed voltage Vref
0.15
Time [s]
0.2
PI boost duty cycle d.
Fig.6: PI controlled boost for changing R and fix Vref.
Fig.6(a, b, c, d) show PI controlled boost system
responses of the voltage V0, current I0, error E and duty
cycle d for changing R and fix Vref. We first simulate the
controlled boost system for a load changing of 30%R, then
60%R at t=0.1s and t=0.2s.
PI boost Output Voltage VO for changing R & fix Vref
60
50
Vout (V)
40
1) Robustness tests
For robustness tests of the fuzzy boost controller, we
study apart the changing influence of load R and reference
voltage Vref. We first simulate the system for a fixed
reference voltage Vref and a load changing of 30%R, then
60%R at t=0.1s and t=0.2s. After that, we change the
reference voltage Vref while the load R is fixed.
a) Changing load R and fixed voltage Vref
30
20
10
0
0
B. Simulation of fuzzy controlled boost converter
Fig.7(a, b, c, d) show fuzzy controlled system
responses of the voltage V0, current I0, error E and duty
cycle d for fix Vref and changing R.
PI V0
Vref
0.05
0.1
0.15
Time [s]
0.2
0.25
0.3
Fuzzy boost Output Voltage VO for changing R & fix Vref
60
a)
PI boost Output voltages V0 and Vref.
50
Fuzzy V0 (V)
3
PI boost Output Current IO for changing R & fix Vref
2.5
Iout (A)
2
1.5
1
40
30
Fuzzy V0
Vref
20
10
0.5
0
0
0.05
0.1
0.15
Time [s]
0.2
0.25
0.3
0
0
0.05
a)
b)
PI boost Output current I0.
0.1
0.15
Time [s]
0.2
0.25
Fuzzy boost output voltages V0 and Vref.
0.3
Fuzzy boost Output Current IO for changing R & fix Vref
Fuzzy boost Voltage VO for changing Vref & fix R
60
2.5
50
2
40
Vout (V)
Fuzzy I0(A)
3
1.5
1
Fuzzy V0
Vref
30
20
0.5
10
0
0
0.05
0.1
b)
0.15
Time [s]
0.2
0.25
0.3
0
0
0.15
0.2
0.25
Time [s]
Fuzzy boost output Voltage V0 for changing Vref.
a)
Fuzzy boost output current I0.
Fuzzy boost Error E for changing R & fix Vref
50
0.05
0.1
0.3
Fuzzy boost Current IO for changing Vref & fix R
1.2
Fuzzy I0
1
40
Iout (A)
Error E (V)
0.8
30
20
0.6
0.4
10
0.2
0
-10
0
0
0
0.05
0.1
c)
0.15
Time [s]
0.2
0.25
0.05
0.1
0.2
0.25
0.3
0.3
b)
Fuzzy boost output Current I0 for changing Vref .
Fuzzy boost error.
Fuzzy boost Error E for changing Vref & fix R
60
0.5
0.15
Time [s]
Fuzzy Error E
Fuzzy boost Duty Cycle d for changing R & fix Vref
40
Fuzzy E (V)
0.4
d
0.3
20
0.2
0
0.1
-20
0
0
0
0.05
0.1
d)
0.15
Time [s]
0.2
0.25
0.05
0.3
c)
Fuzzy boost duty cycle d.
0.1
0.15
Time [s]
0.2
0.25
0.3
Fuzzy boost Error E for changing Vref.
Fuzzy boost Duty Cycle d for changing Vref & fix R
0.5
Fuzzy duty cycle d
Fig.7 Fuzzy controlled boost responses for changing R and fix Vref .
0.3
d
Fig.7 shows that the system output voltage V0 pursues the
reference voltage Vref for any load change. We observe a
small pick of 1V at t=0.1s and another one of 2V at t=0.2s
due to the load value changing with very small errors.
0.4
0.2
0.1
b) Changing voltage Vref and fixed load R
Fuzzy controlled system responses for the output
voltage V0, output current I0, error E and duty cycle d for
changing Vref at t= 0.1s and t = 0.2s for Fixed R, are
shown in fig.8(a, b, c, d).
0
0
0.05
d)
0.1
0.15
Time [s]
0.2
0.25
Fuzzy boost duty cycle d for changing Vref.
Fig.8 Fuzzy controlled boost for changing Vref and Fix R.
0.3
C. Simulation of PI controlled boost converter
Fig.9(a, b, c, d) show PI controlled system
responses of the voltage V0, current I0, error E and
duty cycle d for changing Vref and fi
PI boost Voltage VO for changing Vref & fix R
100
PI V0
Vref
60
D. PI and fuzzy controllers Comparison
PI and fuzzy boost controller performance parameters
are compared using their responses in fig.10 showing the
settling time, overshoot and static error. We use 5%
criterion as shown in fig.10 to estimate system speed of
response.
40
60
20
50
0
0
0.05
a)
0.1
0.15
Time [s]
0.2
0.25
0.3
PI boost output Voltage V0 for changing Vref.
30
ts=22.5ms
ts=5ms
Fuzzy Controller
PI Controller
+5%
65%
20
PI boost Current IO for changing Vref & fix R
2
PI & Fuzzy boost Voltages VO for fix Vref & R
40
Vout [V]
Vout,Vref (V)
80
Fig.8 (a, b, c, d) show fuzzy boost controller responses
with better tracking compared to PI boost controller
illustrated in fig.9 (a, b, c, d).
PI Current I0
10
Iout (A)
1.5
0
0
1
0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
Time (s)
Fig.10 Fuzzy and PI boost output voltages.
0.5
0
0
0.05
b)
0.1
0.15
Time [s]
0.2
0.25
0.3
From fuzzy and PI output voltages of fig.10, we illustrate
performance parameters in table.2.
PI boost output Current I0 for changing Vref .
TABLE.2 PI AND FUZZY PERFORMANCE PARAMETERS.
PI boost Error E for changing Vref & fix R
60
Regulators
PI Error E
PI
Fuzzy
40
Settling time
Ts (5%) (ms)
22.5
5
Overshoot
DMax (%)
14.8
3.6
Mean
Error E(V)
0.5623
0.0621
PI E (V)
20
Fig. 10 shows fuzzy output voltage Vo, in steady state,
reaching the reference voltage Vref = 50V in a shorter
setling time ts (fuzzy) =0.005s for a smaller overshoot
Dmax(fuzzy) =3.6% than PI settling time ts(PI)=0.0225s
and a PI overshoot Dmax(PI)=14.8% with better accuracy
EFuzzy=0.0621<EPI=0.5623.
0
-20
-40
0
0.05
0.1
c)
0.15
Time [s]
0.2
0.25
0.3
V.
PI boost Error E for changing Vref.
PI boost Duty Cycle d for changing Vref & fix R
1
PI duty cycle d
0.8
d
0.6
0.4
0.2
0
0
0.05
d)
0.1
0.15
Time [s]
0.2
0.25
PI boost duty cycle d for changing Vref.
Fig.9 PI controlled boost for changing Vref and Fix R.
0.3
CONCLUSION
Various tests have shown that fuzzy regulator is
capable of reducing effect of different disturbances such
as load changes and reference voltage variations common
in industry. Fuzzy regulator yields better dynamic
performance than PI especially in transient state
concerning speed of response and smooth tracking without
overshoot with better steady state accuracy which
confirms the overall effectiveness of fuzzy logic method.
Unlike conventional PI boost controller, fuzzy controllers
are based on artificial intelligence rather than on the plant
model ill-defined leading to greater complexities in the
design of the control system. For this reason, we have
used small signal behaviour for the PI boost controller
modelling. Results show that intelligent control systems
give better performance parameters in such cases and can
provide accurate control over a wide range of operating
conditions. They are able to make approximations and
intelligent guesses in order to come out with good results
under given set of constraints for any type of dc-dc
structures. To improve performance, we will combine
fuzzy logic with GA or PSO.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
Y. Song, A.T Johns, “Applications of Fuzzy Logic in Power
Systems, Part 1,” IEE Power Engineering Journal, Oct. 1997.
Y. Song, A.T. Johns, “Applications of Fuzzy Logic in Power
Systems, Part 2,” IEE Power Engineering Journal, pp. 185–190,
Aug. 1998.
C. Cecati, A. Dell’Aquila, A. Lecci, and M. Liserre,
“Implementation issues of a fuzzy-logic-based three-phase active
rectifier employing only voltage sensors,” IEEE Trans. Ind.
Electron., vol. 52, no. 2, pp. 378–385, Apr. 2005.
G. O. Cimuca, C. Saudemont, B. Robyns, M. M. Radulescu,
“Control and performance evaluation of a flywheel energy-storage
system associated to a variable-speed wind generator,” IEEE
Trans. Ind. Electron., Vol. 53, No. 4, pp. 1074–1085, Aug. 2006.
S. Chakraborty, M. D. Weiss, M. G. Simões, “Distributed
intelligent energy management system for a single-phase highfrequency ac microgrid,” IEEE Trans. Ind. Electron., Vol. 54, No.
1, pp. 97–109, Feb. 2007.
S. Sanchez-Solano, A. J. Cabrera, I. Baturone, F. J. Moreno-Velo,
M. Brox, “FPGA implementation of embedded fuzzy controllers
for robotic applications,” IEEE Trans. Ind. Electron., Vol. 54, No.
4, pp. 1937–1945, Aug. 2007.
J. Harris, “Fuzzy Logic Applications in Engineering Sciences,”
Springer, 2006.
A. Gad, M. Farooq, “Application of fuzzy logic in engineering
problems,” in Proc. 27th Annu. Conf. IEEE Ind. Electron. Soc.,
Denver, CO, Vol. 3, pp. 2044–2049, Nov. 29–Dec. 2, 2001.
F. Betin, D. Pinchon, A. Jaafari, “Robust speed control of a
stepping motor drive using fuzzy logic,” Proceedings of the 1998
IEEE International Conference on Control Applications Trieste,
Italy, September 1998.
M. Cheng, Q. Sun, and E. Zhou, “New self-tuning fuzzy PI control
of a novel doubly salient permanent-magnet motor drive,” IEEE
Trans. Ind. Electron., Vol. 53, No. 3, pp. 814–821, Jun. 2006.
R.-J. Wai, K.-H. Su, “Adaptive enhanced fuzzy sliding-mode
control for electrical servo drive,” IEEE Trans. Ind. Electron., Vol.
53, No. 2, pp. 569–580, Apr. 2006.
R. Oruganti, M. Palaniapan, “Inductor voltage control of buck-type
single-phase ac-dc Converter,” IEEE Transactions on Power
Electronics, Vol. 15, No. 2, pp.411417, 2000.
Y. Berkovich, A. Loinovici, “Dynamic model of PWM zerovoltage-transition DC-DC boost Converter,” Proc. IEEE
International Symposium on Circuits and Systems, Orlando, FL,
Vol. 5, No. 2, pp.V-254V-260, 1999.
M. Veerachary, “Design of robust digital stabilizing controller for
fourth-order boost DC-DC converter: A quantitative feedback
theory approach,” IEEE Trans. Ind. Electron., Vol. 59, No. 2, pp.
952963, Feb. 2012.
Said El Beid, Said Doubabi, “DSP-Based Implementation of Fuzzy
Output Tracking Control for a Boost Converter,” IEEE Transaction
on Industrial Electronics, Vol. 61, No. 1, Jan. 2014.
B. Rajasekaran, M. L..Bharathi “High-Efficiency Modular High
Step-Up Interleaved DC-DC Boost Converter Using Fuzzy
Controller,” International Journal Of Scientific & Technology
Research, Vol. 3, No. 2, Feb. 2014.
F. Ueno, T. Inoue, I. Oota and M. Sasaki, “Regulation of Cuk
converters Using Fuzzy Controllers,” INTELEC’91. pp.26267,
Nov. 1991.
R. W. Erickson, D. Maksimovic, “Fundamentals of Power
Electronics.” Norwell, MA: Kluwer, 2001.
B. Lehman, R.R Bass, “Extensions of averaging theory for Power
Electronic Systems,” IEEE Transactions on Power Electronics,
Vol. 2, No.4, pp.542553, 1996.
H. Bevrani, M. Abrishamchian, N. Safari-Shad, “Nonlinear and
linear robust control of Switching Power Converters,” Proc. IEEE
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
International conferences on Control Applications, Vol. 1, No.4,
pp.803813, 1999.
K. Goser, Innovation and Application of ANN in Europe
Demonstrated by Kohonen Maps. In Jet Propulsion Laboratory
(NASA) Worshop on “A Decade of Neural Networks, Practical
Applications and Prospects,” JPL, CalTec, Pasadena, pp.9193,
May 1994.
G. W Irwin, K. Warwick, K. J. Hunt, “Neural Network
Applications in Control”, Institution of Electrical Engineers,
London, 1995.
H. B. Verbruggen, H.J. Ziimmermann, R. Bakuska “Fuzzy
algorithm for control,” International Series in intelligent
Technology, Kluwer Academic Publisher, 1999.
Abdul R. Ofoli, Ahmed Rubaai, “Real-Time Implementation of a
Fuzzy Logic Controller for Switch-Mode Power-Stage DC–DC
Converters” IEEE Transactions on Industry Applications, Vol. 42,
No. 6, Nov./Dec. 2006.
Liping Guo, John Y. Hung, R. M. Nelms, “Evaluation of DSPBased PID and Fuzzy Controllers for DC–DC Converters,” IEEE
Transaction on Industrial Electronics, Vol. 56, No. 6, June 2009.
J. Echanobe, I. Del Compo, J. M. Tarela, “Issues concerning the
analysis and implementation of a class of fuzzy controllers,” Fuzzy
Sets and Systems, Vol. 155 pp.252271, 2005.
Leocundo Aguilars, Patricia Melin, Oscar Castillo “Intelligent
Control of a stepping motor drive using a hybrid neuro-fuzzy
ANFIS approach,” Applied Soft Computing, Vol. 3, pp.209219,
2003.
J. J. Buckley, Y. Hayashi, “Fuzzy Neural Network, A survey”,
Fuzzy Sets Syt., Vol. 66, pp.113, 1994.
D. A. Linkens, H. O. Nyongesa “Genetic Algorithms for Fuzzy
Control  Part I and II,” Inst Elect. Eng. proc. Contr., Theory
Applicat., Vol. 142, pp.161176, 1995.
W. Pedryez, “Fuzzy Control and Fuzzy Systems,” New York,
Wiley, 1989.
C. C. Lee, “Fuzzy logic in control systems: Fuzzy logic controllerPart 1,” IEEE transactions on systems, man and cybernetics, Vol.
20, No 2, April 1990.
C. C. Lee, “New Defuzzification method for fuzzy control of
power converters,” Industry Applications Conference, Conference
Record of the 2000 IEEE, Vol. 2, pp.12021209, 2000.
J. Kim, B. P. Zeigler, “Designing fuzzy Logic controllers using a
multiresolutional search paradigm,” IEEE Trans. Fuzzy Syst., Vol.
4, pp.213226, Aug. 1996.