Fuzzy Sag

Fuzzy logic control: A
successful example
Standards
Certification
Education & Training
Publishing
Conferences & Exhibits
Copyright 2012 by ISA, www.isa.org
Presented at ISA Automation Week 2012, 24-27 September 2012
Orange County Convention Center, West Concourse, Orlando, Florida USA
Presenter
Michel Ruel, P.E.,
founder and president of TOP Control Inc., now a member of BBA Inc. (900 p)
registered professional engineer, university lecturer, and author of several publications and
books on instrumentation and control.
for over 37 years, he has been solving unusual process control problems in several fields in
more than 16 countries
graduated from Laval University, Quebec Canada, with a Bachelor of Science, Electrical
Engineering (Process and Automation).
member of the following organizations:
ISA, Fellow (International Society for Automation);
IEEE (Institute of Electrical and Electronic Engineers)
PAPTAC (Pulp and Paper Technical Association of Canada)
TAPPI (Technical Association of the Pulp and Paper Industry)
AIChE(American Institute of Chemical Engineers)
PEO (Professional Engineers of Ontario)
OIQ (Ordre des ingnieurs du Qubec)

Agenda
SAG Mill Process and Process Control
Advanced Process Control
Decision Tree
Comparison
Fuzzy Logic Control

Controller Design
Commissioning and Optimization
Results

Process
Measurements (controlled):
Pulp Density
Power
Weight (bearing pressure)
Recirculation Flow
Disturbances
Ore size
Ore Hardness
Crusher Opening
Manipulated
Speed
Feed (tonnage)
Water Flow

Need for Advanced Process Control

Use APC to :
Improve performance,
Stabilize production,
Handle constraints,
Handle interactions,
Protect equipment,
Manage grade changes.
Approaches
Advanced Regulatory Control (PID control +++)

Model Predictive Control
Fuzzy Logic Control
Neural Network
SISO and MIMO

SISO, Single Input, Single Output
Each loop is alone
One model per loop
MIMO, Multiple Inputs, Multiple Outputs

Models for input/output +
Models for interaction +
Models for distrubances

Multi Loop Process Control, MIMO

Disturbances
SP
CO
Controller
PV
Process

Modeling
Small SP excitation (Closed Loop)
Automated Standardized Tests
Normal Operation
Multi Loop
No need to stabilize the process
Automated
Results
Models matrix
Quality of models
Error bound

Agenda
Decision Tree
Comparison
Fuzzy Logic Control

Controller Design
Results

Models are identified or calculated
1-SISO
1 PV, 1 CO
Good process model
Non linear process model or models
scheduling (or PID)
PID controller
PID controller + gain
2-MIMO
n PVs, mCOs
Good process models, weak interactions
Good process models, interactions
Good process models, strong interactions
PID controllers
PID controllers, tuned at
different speeds
PID controllers + decouplers
MPC
PID controllers + decouplers
MPC

10
No process models identified

nor calculated
1-SISO
1 PV, 1 CO
No process model
parameters
PID controller, relaxed tuning
PID controller + logic + enhanced

functions
Best operator can be modeled
Fuzzy logic controller
2-MIMO
n PVs, mCOs
PID controllers,
relaxed tuning parameters
+ logic + enhanced functions
Best operator can be modeled Fuzzy logic controller
Best operator cannot be modeled
No process model
Data available
Neural network
Presented at ISA Automation Week 2012,
September 2012
No data Orange
24-27
Re-design!
County Convention Center, West Concourse, Orlando, Florida USA
11
PID vs. APC

Is Advanced
Regulatory Control
Sufficient?
Yes
No
PID
Can
Process be
Modeled?
Feedforward
Decoupling
Adaptive Gains
Characterizers
Yes
No
MPC
Can Best
Operator
Control?
Yes
Fuzzy Logic

No
Neural Network
12
Comparison
Approach
Model
Rules
Historical
Control
ARC
MPC
FLC
Neural Network
Description
PID, Control
strategies
Process is
modelled
Operator is
modelled
Black box
approach
Usage
Few variables
Good models
Best operator
Based on
historical data
Development
Simple
Moderate
Complex
Black box
Commissioning
Simple
Moderate
Long but easy
Black box
Optimization
Simple
Part of design
Cumbersome
No need
Re-train
Process changes
Simple
Re-model
Review rules and

membership
functions
Maintenance
Simple
Needs expert
Easy
Re-design
Cost
Low
High
Moderate
High

13
Agenda
Decision Tree
Comparison
Fuzzy Logic Control

Controller Design
Results

14
Fuzzy vs. PID
The density
is too high. Ill increase
water flow.
de
1
OP = K p e + edt + Td + Bias
dt
Ti 0
DC
15
Fuzzy Control
Power
Feed
Weight
Speed
IF (power IS high) AND

(Weight IS low) THEN
(Speed medium), (Feed low),
What do high, medium and low actually mean?
These vague, subjective classifications put the fuzzy into fuzzy logic.

16
Crisp Sets vs Fuzzy Sets
22.5 oC
22.5 oC
1
1
cold
16 oC
warm
22 oC
cold
hot
28 oC
Partly warm
26 oC
warm
hot
22 oC
28 oC
Partly cold
Membership function defines:

Membership function defines if
The degree of membership (or fulfillment)
element is or is not a member of the
of any element in the fuzzy set
set, either
in degree of membership = 1.0
out degree membership = 0.0
Partial membership is allowed

22.5 oC is partly (20%) cold and partly
(80%) warm

17
Controller Structure (PLC function blocks)
Inputs)
Fuzzification
Decisions
Defuzzification
(rules)
If T is
AND THEN
(423)
Inputs
Outputs)
Outputs
Xn and dXn/dt
Numerical
Membership function
3 to 7
Adjustable weights
Mix of all fired rules

Membership function
Numerical
3 to 7
Adjustable weights
Rules
Logic (Inputs, Outputs)
Adjustable weights

18
Agenda
Decision Tree
Comparison
Fuzzy Logic Control

Controller Design
Results

19
Controller Structure
Fuzzification
Load (5)
dLoad/dt (5)
OreSize (3)
dOreSize/dt (3)
Recirculation (5)
dRecirculation/dt (5)
Power (5)
dPower/dt (5)
Density (5)
dDensity/dt (5)
Defuzzification
Tonnage (7)
Water flow (5)
Rotation speed(5)
(rules)
If T is
AND THEN
(423)
Rule
shapes
Decisions
weights
Pressure
Pressure
Rate
Power
Power Rate
Tonnage
HH
HH
HH
H
H
L/LL
HH/H
OK
L
OK/L
LL
OK
H/OK/L/LL
OK/L
L/LL
HH/H
L
OK
LLL
OK
123
124
125
126

20
Controller Design
Objectives
to reduce power consumption per ton of ore,
increase throughput,
protect linings and stabilize quality and operation
Determining Rules
Design of experiments (DOE) to determine how the SAG mill
should be operated.
These tests were conducted in different conditions.
All tests were conducted during the summer of 2011
Which resulted in hundreds of rules.
Rules were then chosen to reach the selected goals and to push
the feed rate to its maximum.

21
Controller Design
Membership Functions
Shapes based on DOE
Number based on expected ranges and rules
Rules
More than 500
MIMO
Structured
Programming
PLC fuzzy functions
Workarounds for bugs and optimization

22
Agenda
Decision Tree
Comparison
Fuzzy Logic Control

Controller Design
Results

23

Advisory mode 3 days,
4days, FLC was used during the day
membership functions, shapes and ranges were modified.
rules were also modified while others were added
8th day, FLC was used continuously

Every week,
metallurgists validate the rules and make slight adjustments.
Training
operators, metallurgists, maintenance technicians and engine
Maintenance
Plant personnel maintain the system, modify the controller, add
rules and optimize the controller.
24
Tools
Statisticals to support metallurgists
Historian
Rules used (% time, strength, etc.)
Statistical data on rules and inputs
Key performance indices:
Tons/d, kW/ton, average error, etc.
Performance Monitoring Software

KPI
Utilization
Performance

25
Agenda
Decision Tree
Comparison
Fuzzy Logic Control

Controller Design
Results

26
Results
Utilization > 98%
Commissioning and
Optimization

27
Conclusions
This project was carried out over six months.
The team consisted of:
consultant personnel, metallurgists from the plant and operators.
Operators have quickly gained confidence and

performances have been improved:
Utilization > 98%

Energy per ton has been reduced by 8%
Tonnage per day has been increased by 14%
A production record was achieved during the first week.
The savings generated by the fuzzy logic controller covered the
projects cost in less than three months.

28
Thank you!
Questions?
Michel Ruel, P.E.
Top Control is now Member of BBA Inc.
Department Manager, Optimization and Advanced Control
(877)867-6473 office
(418)569-8543 cell
michel.ruel@bba.ca
http://www.BBA.ca

Fuzzy Sag

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Fuzzy Sag

Caricato da

Copyright:

Formati disponibili

Fuzzy logic control: A

Michel Ruel, P.E.,

Copyright 2012 by ISA, www.isa.org

Fuzzy Logic Control

Copyright 2012 by ISA, www.isa.org

Copyright 2012 by ISA, www.isa.org

Need for Advanced Process Control

Advanced Regulatory Control (PID control +++)

SISO and MIMO

MIMO, Multiple Inputs, Multiple Outputs

Copyright 2012 by ISA, www.isa.org

Multi Loop Process Control, MIMO

Copyright 2012 by ISA, www.isa.org

Copyright 2012 by ISA, www.isa.org

Fuzzy Logic Control

Copyright 2012 by ISA, www.isa.org

Models are identified or calculated

Good process models, strong interactions

Copyright 2012 by ISA, www.isa.org

No process models identified

PID controller, relaxed tuning

PID controller + logic + enhanced

PID vs. APC

Copyright 2012 by ISA, www.isa.org

Long but easy

Review rules and

Copyright 2012 by ISA, www.isa.org

Fuzzy Logic Control

Copyright 2012 by ISA, www.isa.org

Fuzzy vs. PID

IF (power IS high) AND

Copyright 2012 by ISA, www.isa.org

Crisp Sets vs Fuzzy Sets

Membership function defines:

Partial membership is allowed

Copyright 2012 by ISA, www.isa.org

Controller Structure (PLC function blocks)

Mix of all fired rules

Copyright 2012 by ISA, www.isa.org

Fuzzy Logic Control

Copyright 2012 by ISA, www.isa.org

Copyright 2012 by ISA, www.isa.org

Copyright 2012 by ISA, www.isa.org

Copyright 2012 by ISA, www.isa.org

Fuzzy Logic Control

Copyright 2012 by ISA, www.isa.org

Commissioning and Optimization

8th day, FLC was used continuously

Performance Monitoring Software

Copyright 2012 by ISA, www.isa.org

Fuzzy Logic Control

Copyright 2012 by ISA, www.isa.org

Copyright 2012 by ISA, www.isa.org

Operators have quickly gained confidence and

Utilization > 98%

Copyright 2012 by ISA, www.isa.org

Potrebbero piacerti anche