Introduction to Neural Networks and Fuzzy Logic Lecture

Introduction to Neural Networks and Fuzzy Logic Lecture 11 Dr. -Ing. Erwin Sitompul President University http: //zitompul. wordpress. com 2 0 1 8 President University Erwin Sitompul NNFL 11/1

Fuzzy Logic Fuzzy Control Solution: Homework 10 v. small perfect big 1 v. big declining growing constant 1 0. 75 0. 6 0. 4 0. 25 0 5 10 15 20 25 distance to next car [m] – 10 – 5 0 5 10 2 speed change [m/s ] 2. 5 m/s 2 13 m –big –small zero +small +big 1 – 2 – 1 0 1 2 acceleration adj. [m/s 2] President University Erwin Sitompul NNFL 11/2

Fuzzy Logic Fuzzy Control Solution: Homework 10 (Cont. ) 0 0. 4 n Rule 1: IF distance is small AND speed is declining, THEN maintain acceleration. 0 0. 75 0. 4 Rule 2: IF distance is small AND speed is constant, THEN acceleration adjustment negative small. 0. 4 0 0. 6 Rule 3: IF distance is perfect AND speed is declining, THEN acceleration adjustment positive small. 0 0. 75 0. 6 Rule 4: IF distance is perfect AND speed is constant, THEN maintain acceleration. 0. 6 FL-Operators: AND Min OR Max President University Erwin Sitompul NNFL 11/3

Fuzzy Logic Fuzzy Control Solution: Homework 10 (Cont. ) –big 0. 6 –small zero +small +big 1 0. 4 – 2 – 1 0 1 2 acceleration adj. [m/s 2] 1 A 2 A 1 – 2 – 1 0 1 2 acceleration change [m/s 2] President University Erwin Sitompul NNFL 11/4

Fuzzy Logic Fuzzy Control Solution: Homework 10 (Cont. ) Output MFs Input 2 MFs Input 1 MFs Finding Solution in Matlab Rule Viewer President University Erwin Sitompul NNFL 11/5

Fuzzy Logic Fuzzy Control Fuzzy Logic Toolbox in Matlab n The toolbox can be opened by typing “fuzzy” in Matlab Workspace n Some variables must be defined: n Number of inputs and outputs n Membership functions of each input and output n Fuzzy rules that will connect the membership functions n Fuzzy set operators, inference core, accumulation, and defuzzification President University Erwin Sitompul NNFL 11/6

Fuzzy Logic Fuzzy Control Fuzzy Logic Toolbox in Matlab n And and OR method pairs: n Min-max n Prod-probor (algebraic product/sum) n Implication: n Min (clipping) n Prod (scaling) n Aggregation : max (accumulation). n Defuzzification : centroid (center of gravity). President University Erwin Sitompul NNFL 11/7

Fuzzy Logic Fuzzy Control Fuzzy Logic Toolbox in Matlab n Now, we utilize the fuzzy toolbox to analyze the inputoutput behavior of the fuzzy control. n Later, the resulting fuzzy control can be applied to control dynamic systems in Simulink environment. n In each session, remember to save and re-open the controller that has been designed: n Save files using Files >> Export >> To Disk n Open files using Files >> Import >> From Disk President University Erwin Sitompul NNFL 11/8

Fuzzy Logic Fuzzy Control Homework 11 A n Read the manual of Fuzzy Logic Toolbox carefully. n Learn how to use the toolbox and get familiar with it. n Redo the HW 10 A in Matlab using Fuzzy Logic Toolbox. Print and submit the related screenshots (hardcopy) and the *. fis file (softcopy) of the fuzzy control. n Submission must be in hardcopy and softcopy. n Incomplete submission will not be graded. n Deadline: Sunday, 1 April 2018. President University Erwin Sitompul NNFL 11/9