09.05.

2016

Fuzzy Control Systems

Lecture 10

Fuzzy Control Systems

 Control applications are the kinds of problems for
which fuzzy logic has had the greatest success and
acclaim. Many of the consumer products that we use
today involve fuzzy control.

Sendai Subway

1
 09.05.2016

Fuzzy Control Systems

 Control systems abound in our everyday life.
 For example, economic systems are large, global
systems that can be controlled; ecosystems are
large, amorphous, and long-term systems that can
be controlled.
 Systems that can be controlled have three key
features: inputs, outputs, and control parameters
(or actions) which are used to perturb the system
into some desirable state.

Fuzzy Control Systems

 A control system for a physical system is an
arrangement of hardware components designed to
alter, to regulate, or to command, through a control
action, another physical system so that it exhibits
certain desired characteristics or behaviour.
 Physical control systems are typically of two types:
open-loop control systems, in which the control
action is independent of the physical system output,
and closed-loop control systems (also known as
feedback control systems), in which the control
action depends on the physical system output

2
 09.05.2016

Fuzzy Control Systems

 Examples of open-loop control systems are a
toaster, in which the amount of heat is set by a
human.

 Examples of feedback control are a room

temperature thermostat, which senses room
temperature and activates the heating or cooling unit
when a certain threshold temperature is reached.

Fuzzy Control Systems

 In order to control any physical variable, we must
first measure it. The system for measurement of the
controlled signal is called a sensor.
 The physical system under control is called a plant.
 In a closed-loop control system, certain forcing
signals of the system (the inputs) are determined by
the responses of the system (the outputs).
 To obtain, satisfactory responses and characteristics
for the closed-loop control system, it is necessary to
connect an additional system, known as a
compensator, or a controller, into the loop.

3
 09.05.2016

Fuzzy Control Systems

Sytem Manipulated Sytem

Compensator, output
input Error variable
or Plant
controller

Sensor

A closed-loop control system

Fuzzy Control Systems

 The output, or response, of the physical system

under control (plant) is adjusted as required by the
error signal.

 Th error signal
The i l e(t)
(t) is
i the
th difference
diff b t
between th
the
actual response of the plant, as measured by the
sensor, and the desired response, as specified by a
reference input.

4
 09.05.2016

Control System Design Problem

Control System Design Problem

5
 09.05.2016

Control System Design Problem

Control System Design Problem

6
 09.05.2016

Assumptions in a Fuzzy Control

System Design

1) The plant is observable and controllable.

2) There exists a body of knowledge.
3) A solution exists.
4) The control engineer is looking for a «good
enough» solution, not necessarily the optimum one.
5) The controller will be designed within an acceptable
range of precision.
6) The problems of stability and optimality are not
addressed explicitly; such issues are still open
problems in fuzzy controller design.

Simple Fuzzy Logic Controllers

 First-generation (nonadaptive) simple fuzzy

controllers can generally be depicted by a block
diagram such as that shown below.

7
 09.05.2016

Simple Fuzzy
Logic Controllers

 The steps in designing a simple fuzzy control system

are as follows:
1) Identify the variables (inputs, states, and
outputs) of the plant.
2) Partition the universe of discourse or the
interval spanned by each variable into a number
of fuzzy subsets, assigning each a linguistic
label (subsets include all the elements in the
universe).
3) Assign or determine a membership function for
each fuzzy subset.

Simple Fuzzy
Logic Controllers

4) Assign the fuzzy relationships between the

inputs’ or states’ fuzzy subsets on the one
hand and the outputs’ fuzzy subsets on the
other hand, thus forming the rule-base.
5)) Choose appropriate
pp p scaling
g factors for the
input and output variables in order to
normalize the variables to the [0, 1] or the
[−1, 1] interval.
6) Fuzzify the inputs to the controller.

8
 09.05.2016

Simple Fuzzy
Logic Controllers

7) Use fuzzy approximate reasoning to infer

the output contributed from each rule.
8) Aggregate the fuzzy outputs recommended
by each rule.
9) Apply defuzzification to form a crisp output.
output

Simple Fuzzy Logic Controllers

 A fuzzy control mechanism includes

 A set of input data
 A set of rules that represents the policies and
heuristic.

 Strategies of the expert decision maker

 A method for evaluation of proposed action
 A method for generating promising actions.

9
 09.05.2016

Fuzzy Control System Design

Example: Pressure Control

Suppose an industrial process output is given in
terms of the pressure.
 We can calculate the difference between the desired
pressure and the output pressure, called the
pressure error (e), and we can calculate the
difference between the desired rate of change of the
pressure, dp/dt, and the actual pressure rate, called
the pressure error rate, (ė).

Example: Pressure Control

Also, assume that knowledge can be expressed in the form
of such IF–THEN rules (just one rule as an example):

 The linguistic variables defining the pressure error: “PB”

and “PM,” the pressure error rate: “NS” and heat input
change: “NM,” are fuzzy, but the measurements of both
the pressure and pressure rate as well as the control
value for the heat (the control variable) ultimately applied
to the system are precise (crisp).

10
 09.05.2016

Example: Pressure Control

Typical closed-loop fuzzy control situation for this example

can be given as:

Example: Pressure Control

This example is illustrated in four steps.

Step 1. Value assignment for the fuzzy input and output
variables:
We will let the error (e) be defined by eight linguistic variables,
labeled A1,A2, . . . ,A8, partitioned on the error space of [−em,+em],
and the error rate ( ė,ė or de/dt ) be defined by seven variables,
variables
labeled B1, B2, . . . , B7, partitioned on the error rate space of [−ėm,
ėm]. We will normalize these ranges to the same interval [−a,+a]
as
Both inputs
are defined
in [−a,+a]

11
 09.05.2016

Example: Pressure Control

MFs for INPUTS:

–For the error, e, the eight fuzzy variables: NB, NM, NS, N0, P0,
PS, PM, PB.
–For the error rate, ė, the seven fuzzy variables: NB, NM, NS, 0,
PS, PM, PB.
–The membership functions for these quantities will be on the
range [−a, a], and these can be shown in Tables shortly for
discrete variables or in figures for continuous variables
(for this examples input-1 can be shown with x = e and input-2 can
be shown with y = ė).

Example: Pressure Control

MF Table for input-1:

12
 09.05.2016

Example: Pressure Control

MF Table for input-2:

Example: Pressure Control

MFs for the OUTPUT

The fuzzy output variable, the control quantity (z), will use seven
fuzzy variables on the normalized universe, z = {−7,−6,−5, . . . ,+7}:

13
 09.05.2016

Example: Pressure Control

Step 2. Summary of control rules:

According to human operator experience, control rules are of the
form:

The complete IF-THEN rules can be shown in a table, called a

Fuzzy Associative Memory (FAM) Table.

Example: Pressure Control

IF-THEN Rules shown in the following FAM Table:

14
 09.05.2016

Example: Pressure Control

Step 3. Conversion between fuzzy variables and precise

quantities:
From the output of the system we can use an instrument to
measure the error (e) and calculate the error rate (ė), both of
which are precise numbers.
– We can convert pprecise numbers into fuzzyy byy using
g the defined
MFs
– We can convert fuzzy quantities into crisp by using any
defuzzification method mentioned in previous chapters.

Example: Pressure Control

Step 4. Development of control table:

–When the procedures in step 3 are used for all e and all ė, we
obtain a control table as shown in the following table.
–This table now contains precise numerical quantities for use by
the industrial system hardware.

15
 09.05.2016

Example: Pressure Control

Example: Pressure Control

–If the values in the table are plotted, they represent a control
surface.
–The volume under a control surface is proportional to the amount
of energy expended by the controller.
–It can be shown that the fuzzy control surface (Figure 13.4) will
actually fit underneath the crisp control surface (Figure 13.5),
which
hi h iindicates
di t th thatt th
the ffuzzy control
t l expends
d lless energy th
than
the crisp control.
–Fuzzy control methods, such as this one, have been used for
some industrial systems and have achieved significant efficiency
(Mamdani, 1974; Pappas and Mamdani, 1976).

16
 09.05.2016

Example: Pressure Control

Example: Aircraft Landing Control

Problem

17
 09.05.2016

Example: Aircraft Landing Control

Problem

Example: Aircraft Landing Control

Problem

18
 09.05.2016

Example: Aircraft Landing Control

Problem

 Step 1: Define membership functions for state

variables as shown below.

Example: Aircraft Landing Control

Problem

 Step 2: Define a membership function for the control

output, as shown below.

19
 09.05.2016

Example: Aircraft Landing Control

Problem

 Step 3: Define the rules and summarize them in an

FAM table.

Example: Aircraft Landing Control

Problem

 FAM table is used to represent IF-THEN Rules of the

fuzzy system.
 This table is a short representation of the rules.
 IF ‘height is L’ and ‘velocity is DL’ THEN ‘force is Z’.
 IF ‘height is L’ and ‘velocity is DS’ THEN ‘force is DS’.

20
 09.05.2016

Example: Aircraft Landing Control

Problem

 Step 4: Define the initial conditions, and conduct a

simulation for four cycles. Since the task at hand is
to control the aircraft’s vertical descent during
approach and landing, we will start with the aircraft
at an altitude of 1000 feet, with a downward velocity
of −20
20 ft/s. We will use the following equations to
update the state variables for each cycle.

Example: Aircraft Landing Control

Problem

21
 09.05.2016

Example: Aircraft Landing Control

Problem

Example: Aircraft Landing Control

Problem

22
 09.05.2016

Example: Fuzzy Control of

Inverted Pendulum

Example: Fuzzy Control of

Inverted Pendulum

 Inverted Pendulum and Fuzzy Control System

 The dynamic equation of the inverted pendulum

system can be expressed as below.

23
 09.05.2016

Example: Fuzzy Control of

Inverted Pendulum

 Step 1: Define membership functions for inputs.

Example: Fuzzy Control of

Inverted Pendulum

 Step 2: Define a membership function for the control

output.

Membership function for output variable u

24
 09.05.2016

Example: Fuzzy Control of

Inverted Pendulum

 Step 3: Define the rules and summarize them in an

FAM table.

Example: Fuzzy Control of

Inverted Pendulum

 Step 4: The fuzzification method with a Singleton

Fuzzifier, inference system with Mamdani Method,
and defuzzification method with Center Average for
this fuzzy control system as shown below.

25
 09.05.2016

Example: Fuzzy Control of

Inverted Pendulum

Reference : Y. Becerikli, B.K. Celik, ‘’Fuzzy control of inverted pendulum and concept of stability using Java
Application’’,Mathematical and Computer Modeling 46, 2007.

26