22
www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed. 10:15a, 11:30a, 12:45p, 2:00p, 3:30p, and 4:45p Exhibit Hall (3B)

Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

Embed Size (px)

Citation preview

Page 1: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID and Fuzzy Logic Control Systems

John Limroth, Software Engineer

Yiannis Pavlou, Applications Engineer

Tues, 10:15a and 11:30a

Wed. 10:15a, 11:30a, 12:45p, 2:00p, 3:30p, and 4:45p

Exhibit Hall (3B)

Page 2: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

OverviewOverview

Control technology• PID control• Fuzzy logic

National Instruments products for control• LabVIEW RT• PID Control Toolset for LabVIEW

Control technology• PID control• Fuzzy logic

National Instruments products for control• LabVIEW RT• PID Control Toolset for LabVIEW

Page 3: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Control Terminology

Terms: Process variable Setpoint Controller output Plant

Terms: Process variable Setpoint Controller output Plant

Examples: Temperature Desired temperature Heater voltage Furnace

Examples: Temperature Desired temperature Heater voltage Furnace

Page 4: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Control

Page 5: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Parameters Proportional gain – Kc

Integral gain – Kc/Ti

• Ti is the integral time constant or “reset time”

Derivative gain – Kc*Td

• Td is the derivative time constant or “rate time”

Page 6: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Gains – PPID Gains – P

Proportional gain K c – “The Sledgehammer”

• Provides immediate controller response to setpoint change, but PV may not settle exactly on SP using proportional control alone

Proportional gain K c – “The Sledgehammer”

• Provides immediate controller response to setpoint change, but PV may not settle exactly on SP using proportional control alone

Page 7: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Parameter Tuning – P only

Page 8: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Gains – PIPID Gains – PI

Proportional gain Kc – “The Sledgehammer”

• Provides immediate controller response to setpoint change, but PV may not settle exactly on SP using proportional control alone

Integral gain Kc / Ti – Fine tuning

• Integrates the error over time to overcome the offset from Proportional alone such that PV = SP. However, integral action may cause overshoot, oscillation, and/or instability problems

Proportional gain Kc – “The Sledgehammer”

• Provides immediate controller response to setpoint change, but PV may not settle exactly on SP using proportional control alone

Integral gain Kc / Ti – Fine tuning

• Integrates the error over time to overcome the offset from Proportional alone such that PV = SP. However, integral action may cause overshoot, oscillation, and/or instability problems

Page 9: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Parameter Tuning – PI only

Page 10: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Gains – PIDPID Gains – PID

Proportional gain Kc – “The Sledgehammer”• Provides immediate controller response to setpoint change, but PV

may not settle exactly on SP using proportional control alone

Integral gain Kc / Ti – Fine tuning• Integrates the error over time to overcome the offset from

Proportional alone such that PV = SP. However, Integral action may cause overshoot, oscillation and/or instability problems

Derivative gain Kc* Td – Whoa…• Used to put the reigns on PI control to prevent overshoot and

oscillation and to add stability

Proportional gain Kc – “The Sledgehammer”• Provides immediate controller response to setpoint change, but PV

may not settle exactly on SP using proportional control alone

Integral gain Kc / Ti – Fine tuning• Integrates the error over time to overcome the offset from

Proportional alone such that PV = SP. However, Integral action may cause overshoot, oscillation and/or instability problems

Derivative gain Kc* Td – Whoa…• Used to put the reigns on PI control to prevent overshoot and

oscillation and to add stability

Page 11: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Parameter Tuning – PID

Page 12: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

PID Autotuning

Page 13: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Fuzzy Logic DataflowFuzzy Logic Dataflow

Page 14: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Fuzzy Logic Control for LabVIEW

Why is fuzzy logic important? Easy to implement an intuitive control

strategy Better control of non-linear systems

• PID control is linear• Fuzzy control is non-linear

Page 15: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Rule-Based ControlRule-Based Control

Example• “If temperature is high, then heater voltage output

should be low.”

Membership sets• What is meant by “high?”

Example• “If temperature is high, then heater voltage output

should be low.”

Membership sets• What is meant by “high?”

Page 16: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Fuzzy Sets

Boolean (or two-valued) sets:• Members belong to a set – non-members do not• Traditional Boolean values (on/off, 1/0)

Fuzzy Sets:• Partial membership to set allowed• Values along continuum of 0 to 1

Page 17: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Boolean Set – “High Body Temperature”

Temperature

Mem

ber

ship

Page 18: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Fuzzy Set – “High Body Temperature”

Temperature

Mem

ber

ship

Page 19: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Fuzzy Logic Control

Page 20: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Fuzzy Logic and PID Combined

IF … AND … THEN …

IF … AND … THEN …

IF … AND … THEN …

Rule base

• • •

Fuzzification Fuzzy Inference Defuzzification

ProcessFuzzy ControllerSet Point Values

Measured Values

PID

Command Variable

P I D

Page 21: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

Fuzzy Logic Design Software

Page 22: Www.natinst.com PID and Fuzzy Logic Control Systems John Limroth, Software Engineer Yiannis Pavlou, Applications Engineer Tues, 10:15a and 11:30a Wed

www.natinst.comwww.natinst.com

OverviewOverview

Control technology• PID control• Fuzzy logic

National Instruments products for control• LabVIEW RT• PID Control Toolset for LabVIEW

Control technology• PID control• Fuzzy logic

National Instruments products for control• LabVIEW RT• PID Control Toolset for LabVIEW