Content
Introduction History of Aristo Logic and Fuzzy logic Stages of fuzzy modeling An Academic example Conclusion References
Fuzzy Logic
ARSTO LOGIC
FUZZY LOGIC
Fuzzy Logic
Fuzzy Logic
a1
a2
a3
a4
Fuzzy Logic
Fuzzy Logic
Fuzzy Logic
Fuzzy Logic
Fuzzy Logic
Fuzzy Logic
10
Rule Base
AND OR A B = min (A, B) A B = maks (A, B)
Fuzzy Logic
11
Example
The users of the heating system wants Less fuel consumption Easy to use Inexpensive More warranty period
Fuzzy Logic
12
Fuzzy Logic
13
Fuzzy Logic
14
Fuzzy Logic
15
Fuzzy Logic
16
M M M M M
1,0
0,8
0,8
0,6 0,5 0,5 5 0,4
x
c b e a d
Fuzzy Logic
17
We should choose the max mambership value from this set. Best heating system company is D.
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 18
Fuzzy Logic
19
Fuzzy Logic
20
Most comman defuzzification system for this model is the center of gravity method
Fuzzy Logic
21
Fuzzy Logic
22
ADVANTAGES Easy to compute Works well with the other techniques suitable for mathematical analysis
Fuzzy Logic
23
Fuzzy Logic
24
Fuzzy Logic
25
Fuzzy Logic
26
Thesis
Modeling Bid Mark-up of International Construction Projects with Fuzzy Logic
Gen, A., 2012 Uluslararas naat Projelerinde Katk Paynn Bulank Mantk ile Modellenmesi, T Yap letmesi.
Fuzzy Logic
27
The other goal is to create a fuzzy logic model to estimate amount of bid mark-up in the light of the obtained datas.
Fuzzy Logic
28
Fuzzy Logic
29
Factor Titles
61 factors which is contained with the literature survey are divided into 5 titles. 1-Factors associated with the employer 2-Factors associated with the project 3-Factors associated with the firm 4-Factors associated with the bidding period and the contract 5-Factors associated with the economical milieu and the risk
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 30
Survey
16 firms with 39 different project participated to survey. The questionnaire which is generally answered by bidding department managers, is provided a reliable database.
Fuzzy Logic
31
Ranking
The factors that affects amount of bid markup, ranked on an importance scale from 1 to 5 by the company representatives. 1: very low 2: low 3: medium 4: high 5: very high
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 32
Fuzzy Logic
33
Method of Working
5 input data for the fuzzy logic model (Employer Factor, Project Factor, Firm Factor, Bidding Period and Contract Factors, Economical Milieu and Risk Factors) and an output data (Total estimated amount of bid mark-up which is estimated as a percentage of construction cost).
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 35
Method of Working
Mamdani type of fuzzy logic modeling method is used because of easily creating and its compatibility with the human behaviour and senses. MATLAB package program was used for creation of the model.
Fuzzy Logic
36
Method of Working
Fuzzy Logic
Method of Working
Fuzzification - First of all, fuzzification of input and output datas is required in order to create a fuzzy logic model. - The fuzzification comprises the process of transforming crisp values into grades of membership for linguistic terms of fuzzy sets. The membership function is used to associate a grade to each linguistic term.
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 38
Method of Working
Forming fuzzy rule base - The second step is forming a fuzzy rule base. - The fuzzy rule base is formed by rules like IFTHAN which connects input variables to the output.
Fuzzy Logic
39
Method of Working
Fuzzy inference process - The relations that formed in fuzzy rule base between input and output fuzzy sets are collected in this process. - The process of fuzzy inference provides to get one result from the system - The form of the output is determined by all inferences of rules that are contained.
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 40
Method of Working
Defuzzification - Defuzzification is the final step of fuzzy logic modeling which converts fuzzy datas to precise results. - Defuzzification is the process of producing a quantifiable result in fuzzy logic, given fuzzy sets and corresponding membership degrees.
Fuzzy Logic
41
Fuzzification
en (2009) states that personal feelings, logic and experiences are prominent references for assigning degree of membership and membership functions. Its quite helpful to overcome so many problems. Therefore, the assignments was made according to this principle. Input values assigned to membership functions and fuzzificated through expert opinions in this study.
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 42
Fuzzification
In survey study, input datas ranked in 1-5 interval. According to this scale: value 1 is very low, value 2 is low, value 3 is medium, value 4 is high and value 5 is very high.
In model study, 5 levels are handled again for fuzzification of input datas. Input datas are fuzzificated with the same sets.
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 43
Fuzzification
Output variable(bid mark-up) is also fuzzificated into 5 level. Triangle membership function is used at 4 set. Trapeze membership function is used at 1 set.
Fuzzification
Very low set: Full membership (=1) for 10 value. =0 for 17.5 value.
Low set: Parameters of the set are *13.5, 18.75, 24+. Medium set: Parameters of the set are *20.5, 25.75, 31+. High set: Parameters of the set are *27.5, 32.75, 38+. Very high set: Trapeze membership function. Starts from 35 value. =1 from 40 to 50 values.
Fuzzy Logic
46
The distribution of the rules by output fuzzy sets. Gen, A., (2012)
Fuzzy Logic
47
Fuzzy Logic
48
Fuzzy Logic
49
Fuzzy Logic
Comparing the survey with the real system. Gen, A., (2012)
53
Fuzzy Logic
Comparing the survey with the real system. Gen, A., (2012)
54
Fuzzy Logic
Fuzzy Logic
55
Comparison between model estimations and survey datas. Gen, A., (2012)
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 56
Distribution graph of model estimations and survey datas. Gen, A., (2012)
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 57
Distribution of model and survey datas on x=y line. Gen, A., (2012)
Fuzzy Logic
59
Fuzzy Logic
60
Conclusion
Fuzzy logic provides an alternative way to represent linguistic and subjective attributes of the real world in computing. It is compatible with human behavior and human logic. It doesnt need a mathematical model for its applications. The software is easy and economical. Its easy to learn, flexible than other techniques.
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 61
Conclusion
Because of the complex and uncertain charachter of construction sector, fuzzy logic is a better solution for decision-making phase.
Fuzzy Logic
62
References
en, Z., 2009. Bulank Mantk ilkeleri ve Modelleme, 3. Bask, Su Vakf Yaynlar, stanbul, Trkiye Gen,A., 2012, Uluslararas inaat projelerinde katk paynn bulank mantk ile modellenmesi, Yksek Lisans Tezi, stanbul Teknik niversitesi Fen Bilimleri Enstits, stanbul Suba, S., Beyciciolu, A., Emirolu,M., 2008, Hafif betonlarda donat aderans dayanmnn bulank mantk yntemiyle modellenmesi, Yap Eitimi Blm Teknik Eitim Fakltesi Dzce niversitesi, DZCE Sar, M., Murat, Y.., Krabal, M., Fuzzy modeling approach and applications
ITU Graduate School of Science Engineering & Technology Construction Management Fuzzy Logic 63
Thank You.
Fuzzy Logic
64