Advances in PID control for processes with constraints · processes with constraints Prof. Julio...
Transcript of Advances in PID control for processes with constraints · processes with constraints Prof. Julio...
![Page 1: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/1.jpg)
Advances in PID control for processes with constraints
Prof. Julio Elias Normey-Rico
Automation and Systems Department
Federal University of Santa Catarina - Brazil
IFAC WC 2020 – Berlin – July 2018
2020 IFAC WORLD CONGRESS
![Page 2: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/2.jpg)
/26Advances in PID control for processes with constraints 2
Process control
Manipulated variables (MV) are constrained
• Amplitude constraints
• Slew-rate constraints
Process variables (PV) need to be constrained
• Operational limits
• Safety
Most used in industry: PID + Saturation blocks in MV + AWP
Actuator limitations
Process operation
Use new ideas to improve PID AWP performance
![Page 3: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/3.jpg)
/26Advances in PID control for processes with constraints
Agenda
1. Motivating example
2. PID control review and wind-up
3. PID control with AWP – Proposed solution
4. Conclusions
3
CSPS - Constrained SISO Process Simulation tool
http://rodolfoflesch.prof.ufsc.br/cspstool
Simulation
tool
![Page 4: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/4.jpg)
Motivating example
![Page 5: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/5.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints 5
Ex1-Simple model with constraints
Simple model with delay
Several closed-loop simulations:
- PID tuning
- No constraints
- Constraints – no AWP
- Constraints – classical AWP
- Constraints – Improved AWP
Umax = 0.2 Umin =-0.2
Dumax=0.2
Ymax =1
![Page 6: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/6.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints 6
Ex1-Simple model with constraints
PID tuning (without constraints)
PID (with saturations)
Saturation MV limits
PID (with AWP)
PID with improved AWP(includes PVConstraints)
![Page 7: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/7.jpg)
PID tuning
![Page 8: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/8.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Unified PID formulation
PID tuning: low frequency approximation of the Filtered Smith Predictor
Advances in PID control for processes with constraints 8
Process models: FOPDT, IPDT, UFOPDT
PID form: Filtered derivative action
Tuning parameter = closed-loop time constant To (trade-off)
ROBUSTNESSTo PERFORMANCE
![Page 9: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/9.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Discrete controller – proper sample-time selection Ts
Discrete controller
Advances in PID control for processes with constraints 9
Parallel form
![Page 10: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/10.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Discrete controller
Advances in PID control for processes with constraints 10
Compact form
![Page 11: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/11.jpg)
PID control with AWP
![Page 12: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/12.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints 12
Saturation problems
ui(k) only decrease when e(k)
changes its signal
But, as ui(k) >> ur(k), needs some
time to go out from saturation
Causes WIND UP
![Page 13: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/13.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints
Anti Windup techniques
No tuning parameter
13
Incremental algorithm
code update:
![Page 14: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/14.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints
Anti Windup techniques
Tuning parameter
14
Back-calculation
Ud
Ui
Up
AWP action
![Page 15: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/15.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints
Error Recalculation (ER)
Recalculate the error signal at every sample to maintain the consistence between u(k) (computed) and ur(k) (applied)
15
Used in the code to
update the error:
e(k-1)=e*(k)
?
Anti Windup techniques
No tuning parameter
Consider:
![Page 16: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/16.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints
Comparing AWP methods
Simple example using the simulation tool
16
Ex2- Anti Windup techniques
• Simulation for a 0,95 step
• PID tuning with To=0,5
• Only MV saturation Umx=1, Umin =0.
![Page 17: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/17.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints
Improved AWP
17
Typical process constraints consider other variables
Classical AWP works with simple saturation
Solution: To map the constraints to the manipulated variable!
How? → Using some prediction ideas (MPC concepts)
![Page 18: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/18.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints
Improved AWP
Mapping constraints (Max case)
Using model
18
Direct
Depends on past values of control action!
u(k) will affect the future values of y(k) Predictions!!!
![Page 19: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/19.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints
Improved AWP
MODEL
Predictions
Using prediction ideas
19
Horizon
We do not know the future and we are going to apply only u(k) at t=kTs
Process delay
![Page 20: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/20.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Advances in PID control for processes with constraints
Improved AWP
Problem to be solved
20
A set of inequations to obtain u(k)
Final constraint
PID AWP control is then computed using this Usat (Repeat for min condition)
![Page 21: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/21.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Example:
Ex3- Improved AWP
Advances in PID control for processes with constraints 21
Known past values
Max constraints
Known past values
![Page 22: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/22.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Ex3- Improved AWP
Advances in PID control for processes with constraints 22
Past dataMin constraint
![Page 23: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/23.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Ex3- Improved AWP
Advances in PID control for processes with constraints 23
Implementation equations for max and min values:
Apply the AWP scheme for PID control with
Simulation Tool Example
![Page 24: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/24.jpg)
Conclusions
![Page 25: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/25.jpg)
/26
Motivating examplePID tuning
PID control with AWPConclusions
Conclusions
• Using MPC ideas PID performance can be improved for process with constraints
• Improved AWP PID algorithms can be the solution in modern real-time distributed control systems for simple constrained systems
• Simulation tools are very important for analysis and design
Advances in PID control for processes with constraints 25
![Page 26: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/26.jpg)
/26
Thanks!
UFSC
IFAC 2020 organizers
For your attention
Advances in PID control for processes with constraints 26
![Page 27: Advances in PID control for processes with constraints · processes with constraints Prof. Julio Elias Normey-Rico Automation and Systems Department Federal University of Santa Catarina](https://reader034.fdocuments.net/reader034/viewer/2022050210/5f5c970c0d6fc602c976877a/html5/thumbnails/27.jpg)
/26Advances in PID control for processes with constraints 27
ReferencesAstrom, K.J. and Hagglund, T. (1995). PID controllers: theory, design, and tuning.
Camacho, E.F. and Bordons, C. (2013). Model Predictive Control. Springer, London.
da Silva, L.R., Flesch, R.C.C., and Normey-Rico, J.E. (2018). Analysis of anti-windup techniques in PID control of
processes with measurement noise. In 3rd IFAC Conference on Advances in Proportional-Integral- Derivative Control,
948-953. Ghent, Belgium.
da Silva, L.R., Flesch, R.C.C., and Normey-Rico, J.E. (2019). Controlling industrial dead-time systems: When to use a
PID or an advanced controller. ISA Transactions. doi:10.1016/j.isatra.2019.09.008.
Fertik, H.A. and Ross, C.W. (1967). Direct digital control algorithm with anti-windup feature. ISA Transactions, 6(4), 317.
Flesch, R.C., Normey-Rico, J.E., and Flesch, C.A. (2017). A unified anti-windup strategy for SISO discrete deadtime
compensators. Control Engineering Practice, 69, 50-60.
Guzman, J.L., Astrom, K.J., Dormido, S., Hagglund, T., and Piguet, Y. (2006). Interactive learning modules for PID
control. IFAC Proceedings Volumes, 39(6), 7-12.
Hippe, P. (2006). Windup in Control: Its Effects and Their Prevention. Springer Science & Business Media.
Normey-Rico, J.E. and Camacho, E.F. (2007). Control of Dead-time Processes. Springer, London.
Normey-Rico, J.E. and Guzman, J.L. (2013). Unified PID tuning approach for stable, integrative, and unstable dead-time
processes. Industrial & Engineering Chemistry Research, 52, 16811-16819.