Download as pdf or txt
Download as pdf or txt
You are on page 1of 25

HEALTH MONITORING FUZZY

DIAGNOSTIC SYSTEMS

Fuzzy Rule-Based Health Monitoring Expert


Systems
Banana Ripeness Classifier
Total percentage of Yellow color in the image
Total percentage of Brown and Black color in the
image
The membership function representation of Category
output.
Fuzzy Rules for Classification

R1: if Y = FEW AND G = FEW AND B&B = FEW THEN


CATEGORY = “UNRIPE”
R2: if Y = FEW AND G = FEW AND B&B = MED THEN
CATEGORY = “OVERRIPE”
R3: if Y = FEW AND G = FEW AND B&B = MANY THEN
CATEGORY = “OVERRIPE”
R4: if Y = FEW AND G = MED AND B&B = FEW THEN
CATEGORY = “UNRIPE”
R5: if Y = FEW AND G = MED AND B&B = MED THEN
CATEGORY = “UNRIPE”
R6: if Y = FEW AND G = MANY AND B&B = FEW THEN
CATEGORY = “UNRIPE”
R: if Y = MED AND G = FEW AND B&B = FEW
THEN CATEGORY = “RIPE”
R8: if Y = MED AND G = FEW AND B&B = MED
THEN CATEGORY = “RIPE”
R9: if Y = MED AND G = MED AND B&B = FEW
THEN CATEGORY = “RIPE”
R10: if Y = MANY AND G = FEW AND B&B =
FEW THEN CATEGORY = “RIPE”
Sample for Unripe Banana images
Sample for Ripe Banana images
Sample for Overripe Banana images
Fuzzy Logic Controller: Water Heater
Example of a Fuzzy Logic Controller: Water Heater
Objective: Keep the water temperature as close to 800C as possible,
in spite of changes in the water flowing out of the tank, and cold
water flowing into the tank.
Step One: Defining the Inputs and Outputs for the FLC
LevelSense (input) 0 (min. val) 10 (max. val)
HeatKnob (output) 0 (min. val) 10 (max. val)
TempSense (input) 0 (min. val) 125 (ma. val)

Step Two: Fuzzify the inputs


1. Symmetrically distribute the fuzzified values across the
universe of discourse
2. Use an odd number of fuzzy sets for each variable so that some
set is assured to be in the middle. The use of 5 to 7 sets is fairly
typical.
3. Overlap adjacent sets (by 15% to 25% typically).
Fuzzy Membership Functions for LevelSense
Fuzzy Membership Functions for TempSense
Step Three: Set Fuzzy Membership Functions for Outputs

Fuzzy Membership Functions for Output HeatKnob


Step Four: Create a Fuzzy Rule Base

XS S M L XL
XS A Good Very
ALittle
Amount Little
S A Good Very Very
A Lot
Amount Little Little
M A Whole A Good Very
A Lot
Lot Amount Little
L A Whole
A Lot A Lot A Little
Lot
XL A Whole A Good
A Lot A Lot
Lot Amount
Fuzzy Rule-Base for Output HeatKnob
Step Five: Defuzzify the outputs (use center of area method)
Example:
LevelSense=6.75, TempSense=65
Crisp input –LevelSense=6.75
Fuzzy outputs with membership values—
Medium: 0.4; Large:0.6; all others: 0.0
Crisp input –TempSense=65
Fuzzy outputs with membership values—
Medium:0.75; Large: 0.25; all others: 0.0
Rules that fire for the fuzzy inputs:
1. If TempSense is Medium (0.75) and LevelSense=Medium(0.4)
then set HeatKnob to AGoodAmount(0.4)
2. If TempSense is Large(0.25) and LevelSense=Medium(0.4) then
set HeatKnob to VeryLittle(0.25)
3. If TempSense is Medium (0.75) and LevelSense=Large(0.6) then
set HeatKnob to ALot(0.6)
4. If TempSense is Large(0.25) and LevelSense=Large(0.6) then
set HeatKnob to ALittle(0.25)

You might also like