Professional Documents
Culture Documents
MSIM4112 L7 Fuzzy
MSIM4112 L7 Fuzzy
MSIM4112 L7 Fuzzy
1 2
Suggest
IF SOMETHING HAPPENED THEN you should…
in the process
Information System
(Fuzzy Logic)
Workflow
Ensure “Do it right” in each process
3 4
What is Fuzziness?
Fuzziness occurs when the boundary of a piece of information is not
clear-cut
Responsive Replenishment
Age 25, 30, 30 or 40
On the basis of fuzzy set theory, an element belongs to a set with a certain
grade of membership
System
More formally a fuzzy set A in a universe of discourse (the numerical range
of the inputs) U is characterized by a membership function:
A : U [0,1] --- (a)
In this case the elements of the universe of discourse can belong to the
fuzzy set with any value between 0 and 1
This value is called the grade of membership. If an element has a value
close to 1, the degree of membership, or truth value is high
The characteristic function of a fuzzy set is called the membership function,
for it gives the degree of membership for each element of the universe of
discourse
5 6
μyoung (25) = 1
Age Grade of Membership μyoung (30) = 0.8
25 1.0 μyoung (50) = 0
30 0.8
7 8
Development of a Responsive Conceptual Model of an RRS
Replenishment System (RRS)
The RRS is developed for managing the right supply of goods
particularly for the convenience store business
The proposed system helps to leverage the responsiveness of
relevant operations to meet market demands
RRS consists of six stages
1. Situation analysis
Some How much you
2. Knowledge acquisition historical need to reorder
3. Data collection sales data RRS
of
4. Fuzzification specified
5. Fuzzy inference engine product
The expertise knowledge is
6. Defuzzification stored in IF-THEN rules format
Some expertise knowledge The procedures comprising fuzzification to defuzzification can be grouped to
stored in IF-THEN rules format9 form the fuzzy system, which embraces the fuzzy logic principle
10
11 12
Fuzzification Membership Function
It is mainly concerned with the conversion of the Characteristics of the fuzzy set are determined by the
membership function, e.g. for the fuzzy set A:
input data set (from the data collection process) into
A
fuzzy sets. n
A ( xi )
To carry out this fuzzification process, two decisive
factors have to be specified in order to determine i 1 xi
the overall features of the fuzzy sets, i.e.
The equation is the general mathematical expression of fuzzy
1.Universe of discourse subset A of x where x is the whole data set and x is an element
of subset A, and μA(xi) is the membership function of element xi
2.Membership function in the universe of discourse when the support set is a finite set
X = {x1, x2, x3 … xn}
13 14
15 16
4 8 12 16 20 24 t 4 8 12 16 20 24 t 4 8 12 16 20 24 t
(hour) (hour) (hour)
Fuzzy Set Intersections Fuzzy Set Intersections
Calculate Membership value to Calculate Membership value to
E.g. Representing Time(t) in Fuzzy Logic Normal (μN) Relatively Long (μRL)
If the average time of selling all soft drink to customer is 15 hours µ T(t)
µ T (t)
RL
Then, the time is associated with the Fuzzy Set of Normal (μN)
N 1
1
t
4 8 12 16 20 24
15
19 20
Defuzzification
Application Example: Soft Drink
Replenishment in Convenience Store Chains
In convenience store chains, needed items of goods have to
Final stage in the fuzzy system be supplied to the stores in a timely manner
Crisp values or linguistic values can be obtained through The fluctuation of the soft drink sales can be fairly significant
the process of defuzzification. In certain cases, the stores can be completely sold out within 12 hours
Means that the shelves are totally empty in the next 12 hours of the day,
The crisp value is commonly used in various industrial and this situation may cause discontent among customers
and control systems where only exact numerical values On the other hand, items unsold within a 24 hours period will remain on
are needed during actual control operations the shelves, resulting in the occupation of precious space in the stores
The store manager needs to maximize the utilization
Normally, the most suitable method is used subject to of the limited space of the store so as to be more
the conditions of operation. Center of area is the cost-effective
most commonly used techniques due to their simplicity Data for three days’ sales are selected for the study
and ease of use Three kinds of input data and one value of output data
were selected as the input and output of the fuzzy system
21 22
25 26
27 28
Membership Functions (1) Membership Function of μT
Calculate the membership value of t = 17.5, the crisp
According to the situation analysis, the universe of value intersects with the predicates function of RL and L
discourse for
and the resultant membership values are
a) t in fuzzy set T is 0 – 24 hours
μ17.5 = μL(t)|t=17.5 = 0.375
b) p in fuzzy set P is 0 – 84 cans as there are seven
dozen in an order and μ17.5 = μRL(t)|t=17.5 = 0.625
c) c in fuzzy set C is 0 – 84 customers (assuming every
customer buys only one can of soft drink)
The three day’s sales information shown in the 0.625
29 4 8 16 20 30
Average time (Hour)
14 28 56 70 84 p
14 28 56 70
No. of Cans
No. of Customers
31 32
Rule Tables (Blocks) of the Fuzzy
Rules Composition System
Average Time (T)
Rules can be displayed in table format which can easily Average Number
be searched of Customer (C) RL
No. of Cans
37 14 28 56 70
Note: ^ denotes the function of intersection 38
No. of Customers
-100% -75 -50 -25 25 50 75 100% -100% -75 -50 -25 25 50 75 100%
0.375
Rule 1 Rule 2
Inventory change IS slightly decreased Inventory change IS slightly decreased
39 40
Implication Results (2) Implication Results (3)
Inventory Change (IC) Inventory Change (IC)
50 50
-100% -75 -50 -25 25 75 100% -100% -75 -50 -25 25 75 100%
-100% -75 -50 25 50 75 100% 0.375
-100% -75 -50 -25 25 50 75 100% -25
Rule 5 Rule 6
Rule 3 Rule 4
Inventory change IS no change Inventory change IS slightly decreased
Inventory change IS no change Inventory change IS slightly decreased
41 42
Rule 2 15.39 – 25 1 – 384.75 15.39
w j C j Aj
N
Rule 3 12.40 0 1 0 12.40
Y
Inventory
j 1
Rule 4 12.40 – 25 1 – 310 12.40
Change
w j Aj
N
Rule 5 20.38 0 1 0 20.38
j 1
Rule 6 15.39 – 25 1 – 384.75 15.39
w, C and A denote the weight, center of gravity and Rule 7 12.40 0 1 0 12.40
area respectively for each individual implication Rule 8 12.40 0 1 0 12.40
result Sum: -1501.75 117.65
45 46
12.76
model-based approach to fuzzy control and decision
j 1
w j Aj
N
j 1
making. Chichester ; New York: Wiley.
47 48