You are on page 1of 8

Training systems for basic and future oriented education

in natural science and engineering

in 96

Conventional Automatic Control


and
Fuzzy Control
combined in one System!
Summary

What is WinFACT ? configurated in such a way can be used e.g. as a


conventional PID controller, as a fuzzy controller,
WinFACT is a new, modular software system which
as a (fuzzy)adapted controller or as a universal
on the one hand disposes of tools for the analysis,
distance simulator.
synthesis and simulation of conventional control
systems and on the other hand contains components
for the treatment of fuzzy systems. The graphical
What does WinFACT offer ?
user surface WINDOWS guarantees an extremely WinNFACT consists of individual and generally
low training effort while offering a very high oper- independent program modules that can be combined
ating comfort. Numerous program interfaces and as you like and between which data can be easily
data formats enable to communicate with various transmitted via various communication channels.
peripherals, external processes, the user’s own soft-
First of all, the software system contains all compo-
ware products as well as external software. For this
nents which are necessary for the analysis and syn-
reason, WinFACT will be of great interest for the
thesis of conventional control systems, such as:
fields
The identification of linear systems by means of
education and training measured curves of the input and output values.
research The identification algorithm is very robust
against interferences, as e.g. noise due to mea-
development. suring, and can be used for any type of input
Along with the software, a powerful variety of hard- signal.
ware modules based on micro-controller technology Analysis of linear transmission systems by
is available and can be programmed directly out of calculating the step response, Bode’s-diagram,
WinFACT via a serial interface. Depending on the locus curve, root locus curve and pole-zero char-
installed firmware, the hardware which had been acteristic.

Standard - MSR Fuzzy - Control


Linear system
analysis Presentation
Fuzzy - Shell Design of Fuzzy
LISA module PID systems
INGO FLOP FuzzyPID
Design of conven- Identification
tional controller
RESY IDA
Fuzzy adapted C - Code
PID controller generator
Simulation and FaPID FALCO
synthesis state
space system
SUSY Block oriented
simulation
self
BORIS C-Code programmed
routines
generation
Driver
Software

Compiler / Linker
Software

Hardware
RS 232 Interface

Micro 4/6/8/10
A/D - D/A converter card (68HC11 resp. 80C537) User own hardware

P R O C E S S
Coordination of the individual software and hardware components of WINFACT

2
Summary

The synthesis of linear controllers. All conven- each WinFACT module: when selecting the
tional standard elements can be used as control respective menu items a respective short
components. description is displayed in the status line of the
program window.
The simulation and optimization of control sy-
stems. Toolbar help
The toolbar help is an essential facilitation
Beside conventional methods, one crucial point of
especially for WinFACT beginners. When it is
the software concept can be found in the field of new activated, a help window containing an explana-
processes as fuzzy logic and fuzzy control. Win- tion of the button function will automatically be
FACT therefor offers modules which cover all le- displayed if the cursor is located on the button
vels beginning with the performance of fuzzy-logic for more than approx. one second. Thus, it is
experiments via the interactive design of control easy to learn all the functions of the various
systems getting to the synthesis and simulation of buttons. The experienced user can naturally dis-
complex fuzzy-control systems. Moreover, all de- engage this help function.
grees of freedom offered by the fuzzy logic can be
utilized to their full extend:
different types of fuzzy sets
several logic operators and inference mecha-
nisms Toolbar help function

a great number of defuzzying methods


Data ranges of numeric parameters
The design and the analysis of hybrid systems con- When entering numeric values via a respective
sisting of conventional and fuzzy components is input dialog it is often desirable to know the
possible on the basis of block diagrammatic sim- permissible data range beforehand in order to be
ulation. Components, e.g. for the graphical evalua- prevented from possible error messages and
tion of measuring and simulation results based on warnings. Therefore, an automatic range infor-
the WINDOWS-MDI standard, round up the program mation has been integrated: in case of numeric
system. input fields in parameter dialogs it is sufficient
to click the right mouse button once to display a
The program system has the user surface which is
message window declaring the permissible data
typical for WINDOWS. For the user it particularly
range.
manifests in the characteristic menu structure with
the optional selection via keywords or toolbars and
the comfortable input dialogues which almost ex-
clude incorrect inputs.

Help? Help!
WinFACT helps the user with a range of various
on-line help functions:
Extensive help texts
All components contain detailed help texts.
Explanation of menu items
Typical input dialog
All menu options are explained automatically in

3
CONVENTIONAL AUTOMATIC CONTROL ...

WinFACT helps the user in every phase when Design of control system according to the me-
designing a control system beginning with the de- thod using the frequency response characteristic
sign of a model via the analysis of the controlled Design of control system by numeric parameter
system until the design of a controller and the sim- optimization
ulation of a closed circuit loop. Design of state controllers
Model construction and identification
The WinFACT module IDA enables to generate
linear system models (with time delay if desired)
with a computer by means of the measured curves
of process input and output parameters. The model
will be specified in form of a fractionally rational
transfer function whose order will be determined
automatically by the program. Beside the real model
design, the IDA is also suitable for solving problems
of model reduction.

System analysis Design of control systems according to the method using


the frequency response characteristic
For the analysis of linear systems WinFACT offers,
for example, the following options: Simulation
Calculating the step response Some WinFACT modules have an integrated sim-
Calculating the frequency response and illustra- ulation option so that the designed control systems
ting it in form of a Bode’s diagram or a Nyquist can be analysed immediately. For more complex
plot simulations, the simulation module BORIS is very
Calculating the root locus curve suitable.
Calculating pole and zero configurations

Simulation of state space systems with LISA


Analysis of linear systems with LISA
Presentation
Design of a control system The module INGO serves for presenting results in
WinFACT offers several methods for the actual an optically pretentious manner and for subsequent
design of control systems: treatment. It allows the graphical representation of
Design of control system according to setting all WinFACT data types, as e.g.simulation results,
rules (“rules of thumb”) trajectories, Bode’s diagrams, contours or three-di-
mensional characteristic diagrams.

4
... AND FUZZY CONTROL

Design of fuzzy systems


In WinFACT, fuzzy-systems are designed by means
of a comfortable fuzzy-shell. It allows to define the
fuzzy-sets and the control base graphically via the
mouse. The rules can be entered either in form of
tables, matrixes or texts whereas it is possible to
change between the different types of representation
at any time. If desired, the rules can be additionally
supplied with an importance factor. The fuzzy-sets
can be specified as singletons or in form of a triangle
or a trapezium. WinFACT disposes of a large
number of different mechanisms for the inference
and defuzzying processes. Three-dimensional representation of the transfer response

C-code generation
After designing the fuzzy controller it can, in case
of need, be converted into the ANSI-C-code, which
can then be transfered to the respective hardware.
The data types used for this process can be selected
by the user in order to guarantee an optimum adap-
tation to the object hardware.

The WINFACT fuzzy-shell FLOP

The fuzzy-controller can be analysed in different


ways:
by single-step inference
by calculating characteristics or characteristic
diagrams
by simulation on the basis of external data sets
C-code generating with FALCO

Simulation of fuzzy control systems


The fuzzy controllers which have been designed
with the fuzzy shell can be directly integrated in the
block diagrammatic simulation BORIS. Thus, even
extremely complex hybrid control systems can be
easily simulated. During the simulation process, a
special fuzzy debugger continually shows the user
exactly what is going on inside of the fuzzy control-
ler.
Performance of inference steps

5
BORIS block diagrammatic simulation and more ...

The block diagrammatic simulation system BORIS choise of several numeric integration methods is
is the basic module of WinFACT. Due to its concept, available for the actual simulation process.
it is not only predestinated for an application as a
simulation tool but can also be applied in the fields
of recording and processing of measured data and
control. In contrast to traditional simulation sys-
tems, BORIS allows the direct integration of fuzzy
systems in the simulation process and, thus, is suit-
able specially for applications in the field of fuzzy
logic and fuzzy control.
Super blocks and user-defined blocks

BORIS contains an extensive library of system


blocks. This library can be extended by the user with
his own system blocks. Subsystems can be com-
bined to macro-systems, so-called super blocks,
which can also be nested. Consequently, it is possi-
ble to create reusable modules in an easy and com-
fortable way which can then be applied for the most
distinctive simulation functions. The parameters of
the system blocks are set via comfortable input
masks. The parameters of most system blocks can
Block diagrammatic simulation with BORIS also be changed during the simulation process.

Beyond its application as a pure software tool, The optimization modul is an ideal supplement to
BORIS offers numerous possibilities of connecting the simulation possibilities and enables an automat-
hardware. In connection with A/D-D/A-converter ic numeric optimization of system components (e.g.
boards it is possible to realize a direct process inter- PID controllers) by means of effectiveness crite-
facing which allows to test a control system that has rions which are freely selectable by the user. Pow-
been projected with a PC. A later transfer to the erful methods of parameter optimization on the basis
object hardware can, on the one hand, be achieved of evolution concepts are used for the optimization
via C-sourcecode generation and, on the other hand, process.
via the WinFACT micro-controller system.
The simulation pattern can be simply configurated
by placing system specific icons to any place on the
screen and by connecting them. The integrated
autorouter has the function to position the connect-
ing lines automatically and in the best possible
manner. The virtual worksheet can be scrolled thus
enabling a clear and extremely fast structuring of
even complex systems. A general function gives a
survey of the total pattern of the system whenever
needed. The complexity of the simulation pattern is
only restricted by the memory capacity of the PC. A
Virtual instruments..

6
The BORIS library of system blocks

Input blocks • Final control element with


constant slew rate
• Generator
• File Functional blocks
• Constant
• Logic elements
• Travel characteristic
• Function of one variable
• Controllable sine-wave
generator (VCO) • Function of several variables
• Simulation time • Determination of extreme values
• Signal source • Minimum/maximum
• Statistical functions
Dynamic blocks • Sample-and-hold element
• Controllable sample-and-hold
• P-element
element
• PT1-element
• Analog switch
• PT2-element
• Analog changeover switch
• PT1T2-element
• PTn-element Digital modules
• Limited integrator (I-element)
• Logic circuit with one input
• Resettable limited integrator
• Logic circuit with two inputs
• Derivative unit
• RS-flip-flop
• PID controller
• D-flip-flop
• Adaptive PID controller with
controllable limiting • JK-flip-flop
• All-pass network, first order • Mono-flop
• All-pass network, second order • Forward-backward counter
• Dead-time element • Digital comparator
• Lead/lag element • Analog comparator
• Transfer function G(s) • Zero crossing sensor
• User-defined differential
equation system Control elements
• Basic delay z-1 • Simulation stop
• z-transfer function H(z) • Simulation delay

Static blocks Output blocks


• Limiter • Time characteristic
• Threshold characteristic • Oscillograph
• Dead zone • Analog display
• Two-point characteristic • Digital display
• Three-point characteristic • Bar instrument (bar graph)
• Hysteresis characteristic • Trajectory display
• Three-point element with (x-y graphic chart)
hysteresis • State indication
• Fuzzy controller • FFT
• User-defined characteristic • File (File-Output)
• Table file (Excel formate) A small variety of BORIS
Final controlling elements • Signal depression system block types

• Final control element with


slew rate delimitation Micro-controller blocks

7
Lehrsysteme für die zukunftsorientierte Aus- und Weiterbildung
in Naturwissenschaft und Technik

Liquid level control Ship stabilization


by means of the measuring interface "Comenius-E" by means of the measuring interface "Comenius-E"
and the WinFACT96-Software and the WinFACT96-Software

WinFACT 96 can also be used to link real control loop ele- Equipment 14 00 330
ments, such as controlled systems to the structure of the Measuring interface "Comenius-E"
control loop. The measuring interface "Comenius E" devel- incl. Measuring software under Win-
dows
oped by ELWE is used as the interface between the PC and
the real control loop elements. The connection to the PC is 4 analog inputs, 0 ... ±100 V
established via the RS232 connection cable which is part of 2 analog outputs, 12 bit, 0 ... ±10 V
the delivery. 8 digital inputs, TTL level
The ELWE experimental panel system "Control engineering" 8 digital outputs, TTL level
contains an extensive program or real control loop elements.
The rocker (a position-controlled system with a ball, art. no. P 9.12 10 09 112
10 09 141, not illustrated), the level-controlled system and the Level controlled system
ship stabilizing system are controlled systems that very suit- A closed liquid system with one
able for fuzzy-logic control processes. measuring transducer each for
the measurement of flow and
WinFACT 96, Type A . . . . . . . . . . . . . . . 50 09 002 level.
consisting of BORIS und INGO.

WinFACT 96, Type B . . . . . . . . . . . . . . . 50 09 012 P 9.11 10 09 111


consisting of BORIS, INGO and FLOP. Ship stabilizing system
A model of a ship with two ballast
WinFACT 96, Type C (full version) . . . . . . . 50 09 022 containers (closed liquid system)
consisting of BORIS, INGO, FLOP, LISA, FALCO, SUSY, and sea-wave simulator.
RESY, IDA and FuzzyPID. Two feedback variables.

Demo diskette WinFACT 96 . . . . . . . . . . . 50 09 000


free of charge, in German

For offers and prices please contact:

Factory / Head office: ELWE-Lehrsysteme GmbH • Elwestraße 6 • Tel. (05306) 930-412 • eMail: vt@elwe.com
D - 38162 Cremlingen / Schandelah • Fax (05306) 930-404 • Internet: http://www.elwe.com

Subject matter to change! wp 09 01 20 - 01/03 Printed in Germany

You might also like