See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/3621756

Fuzzy logic application to power electronics and drives – an overview

Conference Paper · December 1995

DOI: 10.1109/IECON.1995.483333 · Source: IEEE Xplore

CITATIONS READS

26 241

2 authors:

G.C.D. Sousa Bimal Bose

Universidade Federal do Espírito Santo University of Tennessee
41 PUBLICATIONS   938 CITATIONS    178 PUBLICATIONS   13,481 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Design of photovoltaic systems View project

All content following this page was uploaded by G.C.D. Sousa on 11 April 2015.

The user has requested enhancement of the downloaded file.

 Fuzzy Logic Applications to Power Electronics and Drives - An Overview

Gilbert0 C. D. Sousa Bimal K. Bose

Departamento de Engenharia Eletrica Electrical and Computer Engineering
Universidade Federal do Espirito Santo The University of Tennessee
c . P. 01-9011 Knoxville, TN 37996 - USA
Vitoria, ES 29060-970 - Brazil

Abstract - Applications of fuzzy logic (FL) to power outperforming other control techniques, such and MRAC and
electronics and drives are on the rise. The paper discusses sliding mode controllers.
some representative applications of FL in the area, In spite of its practical success, fuzzy logic has been
preceded by an interpretative review of fuzzy logic criticized for its limitations, such as the lack of a formal
controller (FLC) theory. A discussion on design and design methodology and the difficulty in predicting stability
implementation aspects is presented, that also considers and robustness of FL controlled systems. In reality, these
the interaction of neural networks and fuzzy logic aspects have evolved considerably in the past few years, as
techniques. Finally, strengths and limitations of FLC are the heuristic approaches, commonly used in FLC design have
considered, including possible applications in the area. gradually been replaced by more formal methods.
In the PE&D field, FL has been applied to various
I - INTRODUCTION problems, such as robust feedback control of a
phase-controlled converter dc drive system [ 11, efficiency
Fuzzy logic (FL), the logic of approximate reasoning, optimization of drive systems [3], hierarchical control of
continues to grow in importance, as its application to a circulating-current converters [5], slip gain tuning of
number of practical problems further demonstrates its vector-controlled induction motor drive system [6], servo
usefulness. In fact, fuzzy logic has been used in areas such as drives [7]-[ lo], induction machine direct self control drives
process control, estimation, identification, diagnostics, [ 111, estimation of power electronic waveforms [ 121, wind
agriculture, medicine, stock market, etc. However, process generation systems [4], ultrasonic motor control [9], etc.
control is by far its most important and visible application. This paper presents an overview of some relevant work in
Lofti Zadeh, father of fuzzy logic, has classified computing the area, preceded by an insightful review of FLCs. Salient
as hard computing and soft computing. The computations design and implementation aspects of FLCs are also
based on Boolean algebra and other crispy numerical considered, along with an analysis of FLC merits and
computations are defined as hard computing, whereas fuzzy Iimitations.
logic, neural network and probabilistic reasoning techniques,
such as genetic algorithm, chaos theory, belief networks and 11-FUZZY LOGIC CONTROLLERS REVISITED
parts of learning theory are categorized as soft computing.
Soft computing differs from conventional (hard) computing Several interpretations of fuzzy logic systems have been
in that, unlike hard computing, it is tolerant of imprecision, proposed. One can view them as special type expert systems
uncertainty and partial truth. Soft computing is more (ES), where the symbolic processing is not as rigorous as in
analogous to thinking of human mind. the conventional ES. On the other extreme, they might be
Fuzzy Logic applications to the control of power seen as pure numeric input/output mapping, quite similar to
electronics and drive (PE&D) systems have been increasing that of an artificial neural network (ANN). In most
exponentially in the past few years. Power converter and applications, however, both aspects are simultaneously
drive systems possess inherent characteristics, such as present, and this ability of providing a mathematical
non-linearities, unavailability of a precise model or its interpretation from a qualitative description is its major
excessive complexity, that make them well suited for FL strength. Some of these aspects will be reviewed here.
control.
The fuzzy logic controller (FLC) of a given process is A . Fuzzy logic controllers as approximators
capable of embedding, in the control strategy, the qualitative
knowledge and experience of an operator or field engineer In many situations FL constitutes an efficient way to
about the process. Therefore, fuzzy logic plays the role of a represent a non-analytical mapping, such that an output value
suitable ”user interface”, in the task of translating designer’s y can be infcrrcd from a given input x. In order to construct
insight about the system into the control law, resulting in an the fuzzy model, the input/output pattern must be known, as
inherently nonlinear adaptive controller, capable of illustrated in Fig. 1.

$4.000 1995 IEEE

0-7803-3026-9/95 57
 Iw

Fig. 2: Linear PI controller mapping.

w not yet zero, but approaches zero at a satisfactory rate. A

X
typical rule base for this class of FLCs is given in Fig. 3, in
Fig. 1: Fuzzy system representation. which the sliding mode is identified by the zero valued
diagonal. For the linear PI controller, such condition is
Membership functions can then be assigned to ranges of characterized by Au(k)=O or:
values for both input and output variables, such that proper
covering of their universes of discourse (domain and range) ce(k) = (-ki / kp)e(k)
is obtained. A certain amount of overlap among the
membership functions is desirable from a robustness
viewpoint. This corresponds to the intersection of the control plane with
A set of fuzzy rules in the form of "IF x is A THEN y is the (e,ce) plane. For the rule base of Fig. 3, when e(k) is not
B" (A and B are fuzzy sets in the universes of discourse of self-correcting, the control increment Au(k) depends on the
x and y, respectively) is obtained from expert knowledge, or sign and magnitude of e&) and ce(k), and its magnitude is
generated by neural or statistical procedures. It has been somewhat proportional to the distance of the (e,ce) cell to the
demonstrated [ 131 that a fuzzy system can approximate any sliding line [ 141. On the other hand, for the conventional PI
continuous function to any degree of accuracy. It can be seen controller, Au@) is a linear function of e@) and ce(k) (1) that
from Fig. 1 that the fuzzy system approximates the unknown is directly related to the distance of the (e(k),ce(k)) point to
function with patches. Higher accuracy can be obtained using the sliding line ( 2 ) . Consequently, the FLC can be viewed as
a larger number of membership functions, at the expenses of a generalization of the linear PI controller, as depicted in Fig.
increased computational cost. 4.

B. The Fuzzy Proportional Integral (PI) Controller helm NM NS ZE PS PM PB

The large majority of applications of FLC lies in this

category. This class of controllers generates the incremental
control signal (Au(k)) from error (e(k)=y*(k)-y(k)) and error
rate (ce(k)=e(k)-e(k-1)). Reasons for such popularity include
the fact that humans are more sensitive to variations in the
data been sensed or actuated upon, than to their actual
values. Furthermore, this type of controller ensures zero
steady-state error for most applications, what can not be
achieved by FLCs that generate the actual control signal (U) Fig. 3: Typical rule base for fuzzy PI controllers.
directly from e(k) and ce(k).
Let us consider a discrete time control system. For a linear
(conventional) PI controller, the incremental control signal is
computed from:
A u(k) = k,ce(k) + kie(k) (1)

This corresponds to a linear mapping from input space (e,ce)

to output space (Au), as illustrated in Fig. 2 . For a PI type
FLC, the rules are designed with a phase plane in mind,
where the corrective term Au(k) forces system dynamics to
annroach that of the slidinn mode. where the svstem error is Fig. 4: Sample fuzzy controller mapping.

58
Superior performance is achieved when the higher degree of
freedom in the FLC design is properly harnessed.

111-FLCS APPLICATION TO POWER ELECTRONICS AND DRIVES

A. Control of a phase-controlled converter dc motor drive [ I ]

While conventional PID controllers can be designed to

give optimum response for a certain operating condition, their
performance deteriorates when parameter variation, non-
linearities and load disturbance effects are present.
Fig. 5 shows the proposed FL based control for a phase-
controlled, separately excited, dc drive system. Fuzzy PI
controllers, similar to those discussed in section 11, are used
in both speed and current loops. For instance, the speed
controller utilizes speed error e@) and change in error c e Q w"a0
to infer the change in control A u Q , by sup-min composition.
-8.5
The actual output (command current I,*) is calculated by
discrete integration of control actions (I,*Q=I,*(k-l)+AuQ).
A second type of fuzzy controller is employed in the
linearization of converter characteristics, under discontinuous
Fig. 6: Linearization of converter characteristics.
conduction. Its principle of operation is illustrated in Fig. 6.
For continuous current conduction, the converter output
B. Fuzzy logic control of a wind generation system[5]
voltage V, varies linearly with cosine of firing angle a.
Linearity is lost when the converter enters the discontinuous
Fuzzy logic can be used to enhance the performance of a
conduction region, where V, becomes a function of both a variable speed cage machine wind generation system, shown
and armature current I,, as indicated by the trajectory BC. in Fig. 7. The generated power is a non-linear function of
Converter linearization can be reestablished by proper wind velocity and generator speed, as indicated in Fig. 8. In
addition of a compensating Am, such that for a fixed a
order to extract the maximum power, a on-line search is
(=80°), the horizontal trajectory BD is obtained. While a implemented, by FLC-1. For a given wind velocity of Vw4,
closed-form equation for Aa does not exist, a fuzzy relation
the output power will initially be at point A, corresponding
can be obtained from the numerical solution, for a limited to a generator speed orl.The FLC 1 will alter the (command)
number of a and I, values. The inherent interpolation
speed in steps Amr, somewhat proportional to observed
capability of FLC assures proper computation of A a for any change in output power AP,, until the actual speed is
given 01 and I, inputs, thus rendering a linear overall
reached, corresponding to maximum output power of point B.
converter characteristics. In this way, faster current loop
response can be obtained for all modes of operation.

"d

I v
-~

Fig. 7: Wind generation system with fuzzy control.

Fig. 5: Fuzzy-controlled dc drive system.

59
 2
A

r. JUMP BY
'
$

Fig. 8: Fuzzy controllers FLC-1 and FLC-2 operation. Fig. 9: Modified slip-recovery drive

The process is restarted if wind velocity changes to Vw2, as il ~ 2 7 0 '

indicated by trajectory BD, until the new maximum power

output is reached (point E). A second fuzzy controller (FLC2)
is used to maximize machine-inverter efficiency, at light load
condition [ 31. Once FLC 1 has driven the generator speed to
its optimum value, such as point E, FLC2 will reduce the
rotor flux by controlling the magnetizing current command
. *
ids in adaptive steps. This method guarantees fast
convergence to optimum flux level, .and is insensitive to
parameter variation and changes in load condition. The output
power is then further increased to point F, as indicated.
A third controller (FLC3) is used to provide robust speed
control against wind vortex and turbine oscillatory torques. It Fig. 10: Drive system phasor diagram.
is a PI type FLC, that infers the change in torque current a) Conventional and b) Modified topology;
reference Aiqs*from speed error and change in error.
A hierarchical fuzzy logic control structure, was selected
to implement the control law. The first level, labeled
"instinct", is responsible for I, control. It is a PI-type fuzzy
C. Adaptive f u z q controller for slip power drives [I 61.
controller that generates the corrective duty cycle increment
Static Scherbius drives posses inherent low power factor, (Ad) from speed error and change in error. It also
due to phase-controlled inverter operation, compounded with incorporates 1, current limiting protection in the rule base.
induction machine VAR requirements. A modified scheme, The second level, that can be likened to the "mechanical
1
shown in Fig. 9, utilizes a boost converter (GTOl), to skills", performs two distinct functions. Firstly, it generates
directly control the dc link current I,, thus freeing the inverter the inverter firing angle, based on external speed reference,
to regulate the line fundamental power factor (displacement actual speed and duty cycle. Secondly, it generates the actual
factor) to unity value, via proper firing angle (a)control. speed reference, required in the first level, by taking into
The principle of operation is illustrated in Fig. 10. In Fig. account starting conditions, acceleration rate and working
10(a) the standard operating mode (with GTOl off) is parameters.
shown, for = 120°, that results in poor line side
The third level is responsible for supervisory control. It
displacement factor (DF). In contrast, Fig. lo@) illustrates monitors relevant system variables, including line DF,
how a unity DF can be achieved, by simultaneously compare actual state with desired condition, and defines the
controlling inverter firing angle (01-225" ) and GTOl's duty corrective actions to be taken by the lower levels.
cycle (d), for the same torque and speed condition. Of course,
IV-FLC DESIGN AND IMPLEMENTATION ASPECTS
forced commutation is required to operate the converter in
such mode.
Typically, the design of a FLC starts with a linguistic
Proper control strategy had to be learned from extensive
description of the control strategy, in the form of IF
simulation study, that revealed a complex interaction of
(condition) THEN (action) rules, that relate some process
control variables (aand d) and output variables (speed, DF,
states (e.g., error and error change) to the appropriate control
I,, etc.) at different system operating conditions.
action. The next step consists in deriving a quantitative

60
description of the linguistic variables, in the form of fuzzy
sets. The design procedure is completed by selecting an
inference method, that defines how the set of control rules
(rule base) is evaluated to yield the control action for a
particular process state. Although the general FLC design
procedure is well known, the fine tuning in most cases has
been performed by heuristic methods, and tends to be time
consuming.
Several FLC design and analysis tools have been
introduced in the past few years, such as Togai's TILSHELL.
By providing a graphical interface for the construction of
membership functions and a rule base editor, the design
procedure can be greatly facilitated. However, PE&D systems
are characterized by possessing linear and non-linear,
continuous and discrete subsystems, with time constants
ranging from microseconds to a few seconds, and
consequently are quite demanding on the FL design and
analysis tools. This explains why most researchers still prefer
to use a general purpose simulation tool (such as MATLAB
or SIMNON) in their work.
Recently, a new version of MATLAB was released that
supports a Fuzzy Logic toolbox. Besides greatly increasing
the flexibility of earlier tools, it also helps to consolidate FL Fig. 11: Neufuz4 block diagram.
as a mainstream control technique.
A typical FL controlled PE&D system will also contain The multilayered neural net, the Fuzzy Rule Generator
conventional control functions, digital filters, plus monitoring (FRG), directly maps weights of different layers into
and protection features. Cost considerations, combined with membership functions and fuzzy rules. To simplify the
the ever increasing capacity of new microprocessors and inference process, the AND operator is interpreted as product,
DSPs, dictates that all control functions must be implemented in the antecedent evaluation (IF portion). The consequents
in a single processor. Therefore, general purpose processors (THEN parts) of the rules also consist of singletons, that are
will be favored, as the solution of choice in the majority of accumulated by the output neuron, to yield the crisp control
applications. Specific fuzzy logic ICs will only be signal.
competitive at the high end applications. As a consequence, The input file should contain enough patters to ensure a
most software tools for FLC design support automatic code good generalization of the neural net, resulting in the best set
generation, typically for C or Assembly language codes, that of rules and membership functions. Therefore, if the
can be readily ported into target processors. An example of simulation is not satisfactory, extra iteration are performed,
such tool is discussed next. after supplying more I/O patterns.
The description of the system using fuzzy logic is much
Neuro-fuzzy tools [I51 cheaper than that using neural nets directly, for real time
control. Furthermore, NeuFuz4 will compile the fuzzy control
The assignment of membership functions and fuzzy rules software directly into COP800 microcontroller assembler
generation are normally dependent on the designer's intuitive codes.
knowledge about the system. As the number of possible rules
increases exponentially with the number of input variables V-MERITS AND LIMITATIONS OF FLCS
and their fuzzy sets, the difficulty in generating the optimum
rule set increases tremendously with system complexity. While it is true that FL control yields faster, more robust
A neural net based automated process to generate the fuzzy control systems, the development time tends to be longer than
rules and membership functions (Neufuz4) has been proposed that of conventional control systems. The basic reason is the
[17], as illustrated in Fig. 11. It requires a numerical intrinsic higher "degree of freedom" of FLC structure, that
input/output description of the system, that is used to learn results in a higher number of parameters to be adjusted.
system behavior, through a modified supervised neural net Another contributing factor is that many application engineers
with error back propagation. Neufuz4 generates bell shaped are not familiar with the new technology. Of course, the use
membership functions, that at the end are approximated by of design tools can help overcome some of these limitations,
trapezoidal curves, better suited for real time computation. as discussed earlier.

61
 FL control has its critics, particularly among the -
VI1 REFERENCES
conventional control theorists. Commonly mentioned
[l]G. C. D.Sousa and B. K. &se, "A fuzzy set theory based control of a
weakness include the lack of a formal design and analysis phase-controlled dc machine drive", IEEE/IAS Annual Meet. Cod. Rec.,
methodologies, including the difficulty in obtaining stability 1991, pp. 854-861.
and robustness indices. While the criticisms is not unfounded, [2] L. A. Zadeh, "Fuzzy Sets", Information and Control, Vol. 8, 1965, pp.
considerable effort has been devoted to solve these problems, 338-353.
with relative success. [3]G. C. D. Sousa, B. K. Bose, J. G. Cleland, R. J. Spiegel and P. J.
Chappell, "Fuzzy logic based on-line eficiency optimization control of an
At the present time, a general design procedure is already indirect vector-controlled induction motor drive", IEEEAECON '93 Cod.
well established. For instance, for the PI type fuzzy Rec., pp. 1168-1174.
controller, the effects of input and output scaling gains on [41M. G. Shoes, B. K. Bose and R J. Spiegel, "Fuzzy logic based intelligent
rising time, overshoot and settling time are well understood. control of a variable speed cage machine wind-generation system", IEEEI
There is also a clear understanding of the desired structure of PESC95 Cod.Rec., pp. 389-395.
[5]L. E. B. da Silva, G. E. April and G. Olivier, "Real time fuzzy adaptive
the rule base, as discussed in section 11. What is left, is the controller for an asymmetrical four quadrant power converter", IEEE/IAS
fine tuning of membership functions and rule base, that is Annual Meet. Cod.Rec, 1987, pp. 872-878.
normally performed via simulation study or experimentation [6]G. C. D. Sousa, B. K. Bose, and Kyung S. Kim, "Fuzzy logic based
with the actual system. This task can also be automated, as on-line MRAC tuning of slip gain for an indirect vector-controlled induction
discussed in [18] for the case of a self-organizing fuzzy motor drive", IEEEAECON '93 Cod. Rec., pp. 1003-1008.
[TY. F. Li and C. C. Lau, "Development of fuzzy algorithms for servo
controller, and more recently, through the use of genetic systems", IEEE Control System Magazine, Vol. 9, no. 3, 1989, pp. 65-72.
algorithm. [Sly. S. Lu and J. S. Chen, "A self-organizing fuzzy sliding-mode controller
It has been also demonstrated [ 171 that a particular class of design for a class of nonlinear servo systems", IEEE Trans. on Ind.
FLC exhibits a behavior similar to that of sliding mode Electronics, Vol. 41, No. 5, Oct. 1994, pp 492-502.
[9]Y. Izuno, R Takeda, and M Nakaoka, "New fuzzy reasoning-based
controllers. Such similarity provides the necessary design
high-performance speed/position servo control schemes incorporating
guidelines, and at the same time assures the robustness of this ultrasonic motor, IEEE Trans. on Ind. Applications, Vol. 28, No. 3, May-June
class of FLC. 1992.
The difficulty in deriving a general, non-conservative [lO]C. Y. Won, S. C. Kim and B. K. Bose, "Robust position control of
stability criterion for FLCs is also present for other non-linear induction motor using fuzzy logic control", IEEEiIAS Annual Meet. Cod.
Rec., 1992, pp. 472-481.
conventional control methods. In fact, most approaches to
[11]S. A. Mir, D. S. Zinger, and M. E. Elbuluk, "Fuzzy controller for inverter
FLC stability analysis are derived from classical non-linear fed induction machines" ,IEEE Trans. on Ind. Applications, Vol. 30, No. 1,
dynamic systems theory. It is normally assumed that a formal Jan./Feb. 1994, pp. 78-84.
model of the plant to be controlled is known, at least [12]M. G. S h o e s and B. K. Bose, "Application of fuzzy logic in the
approximately. estimation of power electronic waveforms", IEEE/IAS Annual Meet. Cod.
Rec., 1993, pp 853-861.
[13]B. Kosko, "Fuzzy systems are universal approximators", Proc. of Fuzzy
-
VI CONCLUSION IEEE, 1992, pp. 1153-1162.
[14]K. L. Tang and R J. Mulholland, "Comparing fuzzy logic with classical
Application of FL to power electronics and drives will controller designs", IEEE Trans. on Systems, Man, and cybernetics, Vol.
continue to flourish in the near future. Most promising SMC-17, NO. 6, Nov./D~c.1987, pp. 1085-1087.
[15]S. Thaler, "Fuzzy rule generation based on a neural network approach",
applications are those where classical control solutions can
Electronic Engineering, July 1993, pp 43-50.
not provide satisfactory performance, typically when the [16]L. E. B. da Silva, B. L. Torres, V. F. da Silva, K. Nakashima, G. E.
model of the plant is not available or highly non-linear, or April, G. Olivier, "Adaptive fuzzy techniques for slip-recovery drive control",
when the system is subject to significant parameter variation. Proc. of Fuzzy IEEE, 1992, pp. 381-388.
FL constitutes a powerful tool to translate human expert [17]D. Driankov, H. Hellendoom, M. Reinfrank, "An introduction to fuzzy
control", Springer-Verlag, 1993.
knowledge into automatic control strategy, a feature that can [18]P. Kosc, G. Griva, "Self-tuning positiodspeed fuzzy logic controllers for
also be exploited in the design of diagnosis and supervisory torque controlled drives", Proc. IPEC-Yokohama, 1995, pp. 1758-65.
control systems. [19]T. Takagi and M. Sugeno, "Fuzzy identification of systems and its
To harness the potential benefits of FL, PE&D engineers applications to modeling and control", IEEE Trans. on Syst., Man and
need to become more familiar with the advancements in FLC Cyber.,Vol. 15, Jan.lFeb., 1985, pp. 116-132.
[20]B. K. Bose, "Expert systems, fuzzy logic, and neural network applications
theory, the capabilities of the new design and analysis tools, in power electronics and motion control", Proceedings of IEEE, 1994, pp
in particular those that take advantage of the growing 1303-1323.
interaction of FL with other emerging technologies, [2l]B. K. Bose, "Fuzzy logic and neural network applications in power
particularly neural networks. A challenge worth taking! electronics", Proc. IPEGYokohama, 1995, pp. 41-45.

62

View publication stats