Fuzzy Inference System
-
Upload
saatnya-yuswanto-menatap-kedepan -
Category
Documents
-
view
200 -
download
2
Transcript of Fuzzy Inference System
FUZZY INFERENCE SYSTEM (FIS) I. Introduction Based on the concept of :
Fuzzy-Rule-Based System, Fuzzy-Expert System, Fuzzy Model, Fuzzy associative memory, Fuzzy Logic Control Fuzzy System Field of application : Automatic control, data classification, decision analysis, expert system, time series prediction, robotics, pattern recognition consists of : 1. Rule base : a selection fuzzy rule 2. Database or dictionary : defines the membership function used in fuzzy rule 3. Reasoning mechanism : perform the inference procedure upon the rule and given facts to derive reasonable output or conclusion input : fuzzy or crisp Output : almost always fuzzy set Sometime we wants a crisp output use defuzzyification to convert fuzzy set into crisp value Block Diagram :
X=(x,y)
Model FIS : 1. Mamdani Fuzzy Models 2. Sugeno Fuzzy Models 3. Tsukamoto Fuzzy Models II. Mamdani Fuzzy Models Block diagram
Defuzzification : 1. Centroid 2. Bisector of Area 3. Mean of Maximum 4. Smallest of Maximum 5. Largest of Maximum There are other Variants of Mamdani Fuzzy Inference System
Tahapan : 1. analisis terhadap permasalahan : - tentukan variable-variabelnya : input dan output - tentukan rentang nilai dari varaibel-variabel tersebut - buat konsep/persepsi untuk setiap variabe 2. Fuzzifikasi : membuat himpunan fuzzy untuk setiap konsep/perspsi tersebut 3. buat rule 4. tentukan metode agregasinya 5. tentukan metode defuzzyfikasi
Rule 1A1 Rule 2A2 X
B1 B2 Y
? Min
ZCOA C1 Max Z?
C1
6. Blok Diagram
III. Sugeno Fuzzy Models Block diagram
Proposed : Takagi, Sugeno and Kang also called : TSK Fuzzy Model Orde 0, linear, orde 2 etc.
IV. Tsukamoto Fuzzy Models Block diagram