FUZZY LOGIC FOR "JUST PLAIN FOLKS "

Chapter 3. Let's Build a Fuzzy Logic Control System

Building a System to Gain Understanding and Familiarity
The easiest and quickest way to understand fuzzy logic control is
to build a fuzzy logic control system; following is one example:
This is a fuzzy logic speed control example, using the same
techniques as used by Professor Mamdani, that you can build for
yourself to get experience with fuzzy logic control. I recommend
you do build some kind of system. I found I began more and more
to understand what fuzzy logic was all about as I tried to make the
system work. The following example system has been reduced in
complexity to make it easier to understand, but the concepts are
the same as those used by Mamdani.
If your application is more demanding than the following example,
you add inputs and "rules"; you do not have to learn new things or
change the approach. In considering this reduced complexity
example, it may be observed that control could have been effected
without going through the fuzzy control exercise we are about to
go through. This would be correct, but only because we are
working with a simple system, only one input and no
discontinuities or aberrations requiring patching.
Following is a system diagram, Figure 3, for a "getting acquainted
with fuzzy" project that provides speed control and regulation for a
DC motor. The motor maintains "set point" speed, controlled by a
stand-alone converter-controller, directed by a BASIC fuzzy logic
control program in a personal computer.

Parts List
(1) IBM or compatible personal computer equipped to run
Microsoft Quick BASIC. IBM is a registered trademark of IBM
Corporation. Microsoft and Quick BASIC are registered
trademarks of Microsoft, Inc.
(2) 8 channel input; 8 bit, analog/digital converter with 8 on-off,
digital output channels and one 8 bit digital/analog output channel.
(3) Signal conditioner (transistor amplifier to adjust levels as
needed)
(4) Transistor - 2N3053.
(5) DC motor, 1.5 V to 3.0 V, 100 ma., 1100 Rpm to 3300 Rpm.
The above speed control system is low cost and suitable for
learning at home or for an application where being rigorously
mathematically correct is not required. The above system
accomplishes the desired purpose and uses triangles to determine
the center of mass (more on this later), but does not add triangles
to compute center of mass as specified by Dr. Bart Kosko (Fuzzy
Thinking, Chapter 10). The fuzzy control action for the above
system is accomplished in a BASIC software program, discussed
below. The controller is a typical low cost ($100) unit, such as the

ANA100 from BSOFT, www.bsof.com. Without the BASIC

program, there is no fuzzy logic action.
National Instruments, www.natinst.com, sells a fuzzy logic system
where the fuzzy control action is accomplished by the software.
National Instruments applications engineers recommend one of
their several analog/digital in, digital/analog out converters for
your application and provide a mathematically correct software
program to produce fuzzy control action. Their system also
provides attractive screen display color graphics. Needless to
say, cost of the National Instruments system is considerably above
the $100 range. One would use the National Instruments
approach for a large, complex system where flexibility and
changes down the road are involved, such as automating a natural
gas liquids plant.
Where using a personal computer is not practical because of
space and weight limitations, fuzzy logic control is also available
utilizing microchips manufactured by Motorola. These microchips
are suitable for fuzzy control applications, www.mcu.motsps.com.
One would use this approach if developing, for example, a fuzzy
logic anti-lock braking system (see "Fuzzy Logic, Revolutionizing
Automotive Engineering; Circuit Cellar INK magazine, November
1997; www.circuitcellar.com).
The steps in building our system are:
1. Determine the control system input. Examples: The
temperature is the input for your home air conditioner control
system. Speed of the car is the input for your cruise control.
In our case, input is the speed in Rpm of the DC motor, for which
we are going to regulate the speed. See Figure 3 above. Speed
error between the speed measured and the target speed of 2,420
Rpm is determined in the program. Speed error may be positive
or negative. We measure the DC output voltage from the
generator. This voltage is proportional to speed. This speedproportional voltage is applied to an analog input channel of our
fuzzy logic controller, where it is measured by the analog to digital
converter and the pesonal computer, including appropriate
software.

2. Determine the control system output. For a home air

conditioner, the output is the opening and closing of the switch
that turns the fan and compressor on and off. For a car's cruise
control, the output is the adjustment of the throttle that causes the
car to return to the target speed.
In our case, we have just one control output. This is the voltage
connected to the input of the transistor controlling the motor. See
Figure 3.
3. Determine the target set point value, for example 70 degrees F
for your home temperature, or 60 Miles per hour for your car.
In our case, the target set point is 2,420 Rpm.
4. Choose word descriptions for the status of input and output.
For the steam engine project, Professor Mamdani used the
following for input:
Positive Big
Positive Medium
Positive Small
Almost No Error
Negative Small
Negative Medium
Negative Big
Our system is much less complicated, so let us select only three
conditions for input:
Input Status Word Descriptions
Too slow
About right
Too fast
And, for output:
Output Action Word Descriptions

Speed up
Not much change needed
Slow down
RULES
Translate the above into plain English rules (called "linguistic"
rules by Dr. Zadeh). These Rules will appear in the BASIC
computer program as "If-Then" statements:
Rule 1: If the motor is running too slow, then speed it up.
Rule 2: If motor speed is about right, then not much change is
needed.
Rule 3: If motor speed is to fast, then slow it down.
The next three steps use a charting technique which will lead to a
computer program. The purpose of the computer program is to
determine the voltage to send to the speed controlled motor. One
function of the charting technique is to determine the "degree of
membership" (see Ch. 1) of the Too slow, About right and Too fast
triangles, for a given speed. Further, the charting technique helps
make the continuous control feedback loop easier to visualize,
program and fine tune.
5. Associate the above inputs and outputs as causes and effect
with a Rules Chart, as in Figure 4, below. The chart is made with
triangles, the use of which will be explained. Triangles are used,
but other shapes, such as bell curves, could also be used.
Triangles work just fine and are easy to work with. Width of the
triangles can vary. Narrow triangles provide tight control when
operating conditions are in their area. Wide triangles provide
looser control. Narrow triangles are usually used in the center, at
the set point (the target speed). For our example, there are three
triangles, as can be seen in Figure 4 (three rules, hence three
triangles).

6. Figure 4 (above) is derived from the previously discussed

Rules and results in the following regarding voltage to the speed
controller:
a. If speed is About right then Not much change needed in
voltage to the speed controller.
b. If speed is Too slow then increase voltage to the speed
controller to Speed up.
c. If speed is Too fast then decrease voltage to the speed
controller to Slow down.

7. Determine the output, that is the voltage that will be sent from
the controller/signal conditioner/transistor to the speed controlled
motor. This calculation is time consuming when done by hand, as
we will do below, but this calculation takes only thousandths of a
second when done by a computer.
Assume something changes in the system causing the speed to
increase from the target speed of 2,420 Rpm to 2,437.4 Rpm, 17.4
Rpm above the 'set point." Action is needed to "pull" the speed
back to 2,420 Rpm. Intuitively we know we need to reduce the
voltage to the motor a little. The "cause" chart and vertical speed
line appear as follows, see Figure 5 below:

The vertical line intersects the About right triangle at .4 and the
Too fast triangle at .3. This is determined by the ratio of sides of
congruent triangles from Plane Geometry:
Intersect point / 1 = 11.6/29 = .4
Intersect point / 1 = 17.4/58 = .3
8. The next step is to draw "effect" (output determining) triangles
with their height "h" determined by the values obtained in Step 7,
above. The triangles to be drawn are determined by the rules in
Step 6. Since the vertical 2,437.4 Rpm speed line does not
intersect the Too slow triangle, we do not draw the Speed up
triangle. We draw the Not much change and the Slow down
triangles because the vertical speed line intersects the About right
and Too fast triangles. These "effect" triangles will be used to

determine controller output, that is the voltage to send to the

speed control transistor. The result is affected by the widths we
have given the triangles and will be calculated. See Figure 6,
below. The Not much change triangle has a height of .4 and the
Slow down triangle has a height of .3, because these were the
intersect points for their matching "cause" triangles; see Figure 4,
above.

The output, as seen in Figure 6 (above), is determined by

calculating the point at which a fulcrum would balance the two
triangles, as follows:
The Area of the Not much change triangle is: 1/2 X Base X Height =
.5 X .04 X .4 = .008. Area of the Slow down triangle is .5 X .08 X .3
= .012.
Compute the controller output voltage by finding the point on the
output voltage, Vdc, axis where the "weight" (area) of the triangles
will balance. Assume all the weight of the Not much change
triangle is at 2.40 Vdc and all the weight of the Slow down triangle
is at 2.36 Vdc. We are looking for the balance point.
Find the position of the controller output voltage (the balance
point) with the following calculation:
(Eq. 1) .008 X D1 = .012 X D2
(D1 is the fulcrum distance from 2.4 V. D2 is the fulcrum distance
from 2.36 V.)

(Eq. 2) D1 + D2 = .04 (from Figure 6)

D1 = .04 - D2
Solving the above by substituting (.04-D2) for D1 in Equation 1
gives D2 = .016 and D1 = .024, therefore the balance point is a
voltage of 2.376 Vdc, and this is the voltage which we have
determined should be applied to return speed to the target value.
See Figure 6, above.
Keep in mind that we are only discussing one sample at one
instant in time, with a resulting controller output voltage; the
controller is sampling several times each second with a resulting
"correction" output following each sample.
The above system was tested with changing loads on the rotating
shaft, and returned the speed of the motor to within 2 % of the
2,420 Rpm set point in less than 1.5 seconds. The accuracy with
which the set point speed can be maintained is determined by the
resolution of the analog to digital and digital to analog conversion
circuits in the fuzzy logic controller. Typical "low cost" resolution
is "8 bit", 256 increments. Higher cost "12 bit" units provide 4,096
increments.
Please note: The above is a very effective, but much simplified,
version of computer based fuzzy logic control systems actually in
use commercially. If your application is of a more demanding,
complex or commercial nature, we suggest you refer to Fuzzy
Thinking, a book by Bart Kosko, Ph.D., Chapter 10, Hyperion, New
York, 1993. Dr. Kosko is one of the world's leading proponents of
fuzzy control and among the most knowledgeable regarding fuzzy
control theory. In the Kosko method, the intersecting triangles are
added, then the total area of the added triangles determined by
integration. Fulcrum location is determined by computer
integration of area "under the curve" to the point of one half the
total area. This sounds complicated, but only requires a few
thousandths of a second for a computer, once the program is set
up.
For more complex systems with additional inputs (for example,
using rate of change as an input in addition to speed error), the
approach is as above, but there are two or more "sub-outputs" to
be considered in arriving at one crisp output to control the system.

This is handled by averaging these sub-outputs with a weighting

determined by the system designer and inserted in the program.
This weighting may be based on theoretical prediction, previous
experience with a similar manual system and/or experimentation
and "tuning" of the system, once it is assembled.
Patch It
For an individual control channel, fuzzy rules cover control
requirements during a certain "range" of operation. In our
example speed control system, one rule covered about right.
There was an actual numerical upper limit and lower limit for about
right. Our control rule for this range is sometimes referred to in
fuzzy logic literature as a "patch." As you can see, the more
patches we have over the control range, the better the control.
Fortunately, most system control problems can be solved with
relatively few patches. A patch, or rule, may be anything that
solves the problem. If the system required it, you could even mix
continuous feedback loop control and off-on control over a
channel's control range, if that solved the problem.
The Program
The fuzzy logic program in the computer directs sending
messages to and receiving messages from the controller, thereby
directing the measurement and control operation and causing
target and actual speed to be displayed. The fuzzy logic controller
receives messages from the computer via BASIC language
commands. Reply messages to the computer from the fuzzy logic
controller are acquired via BASIC.
In this case, the computer was an IBM PC/XT. The programming
language was Microsoft Quick BASIC. The program was compiled
with Microsoft's compiler, but compiling is not essential.
Compiling increases speed of execution and performance. Ideal
computers for fuzzy logic control systems are often ancient IBM
PC-XT computers, available in garage sales for $50. These
computers are of no value for today's software, but work very
adequately for fuzzy logic measurement and control applications.
IBM is a trademark of IBM Corporation. Microsoft and Quick
BASIC are trademarks of Microsoft, Inc.

The portion of the program for the above system system which
examines the input and performs the "triangle" calculations to
arrive at a crisp output follows:
910 IF MS = 2420 THEN MIV = 2.4 : GOTO 5000 'MS-MEASURED
SPEED, MIV-MOTOR INPUT VOLTAGE
920 IF MS < 2420 THEN 2000 ELSE 1000
1000 ' LINES 1010-1110; GREATER THAN 2420 RPM, SLOW DOWN
1010 IF MS > 2449 THEN MIV = 2.36 : GOTO 5000
1020 ' COMPUTE INTERSECT POINT, IPA, FOR 'ABOUT RIGHT'
TRIANGLE
1030 IPA = (2449-MS) / 29
1040 IF IPA =< 0 THEN IPA = .0001
1050 ' COMPUTE INTERSECT POINT, IPS, FOR 'SLOW DOWN'
TRIANGLE
1060 IPS = (MS-2420) / 58
1070 ' COMPUTE MOTOR (TRANSISTOR) INPUT VOLTAGE (MIV)
1080 AAR = .5 * .04 * IPA 'AAR - AREA OF 'ABOUT RIGHT'
TRIANGLE
1090 ASD = .5 * .08 * IPS 'ASD - AREA OF 'SLOW DOWN'
TRIANGLE
1100 D1 = .04 * (ASD / (ASD+AAR))
1110 MIV = 2.4 - D1 : GOTO 5000
2000 ' LINES 2010-2110; LESS THAN 2420 RPM, SPEED UP
2010 IF MS < 2362 THEN MIV = 2.44 : GOTO 5000
2020 ' COMPUTE INTERSECT POINT, IPA, FOR 'ABOUT RIGHT
'TRIANGLE

2030 IPA = (MS-2391) / 29

2040 IF IPA =< 0 THEN IPA = .0001
2050 ' COMPUTE INTERSECT POINT, IPF, FOR 'SPEED UP'
TRIANGLE
2060 IPF = (2420-MS) / 58
2070 ' COMPUTE MOTOR INPUT VOLTAGE (MIV)
2080 AAR = .5 * .04 * IPA 'AAR - AREA OF 'ABOUT RIGHT
'TRIANGLE
2090 ASU = .5 * .08 * IPF 'ASU - AREA OF 'SPEED UP' TRIANGLE
2100 D1 = .04 * (ASU / (ASU+AAR))
2110 MIV = 2.4 + D1
5000 '
The remainder of the program would be determined by the
program requirements of the analog to digital/digital to analog
controller in use. Program statements would be specific to the
hardware selected. Almost any controller should be usable with
the above BASIC statements, so long as the controller could be
programmed in BASIC to measure inputs and send control output
signals. Program execution would cycle in the sequence: 1.
Measure input. 2. Analyze with the fuzzy logic program
statements. 3. Send the output signal.