IA3112 Automatiseringsteknikk Høsten 2018

Slides:

Advertisements

Similar presentations

Chapter 9 PID Tuning Methods.

Advertisements

Tuning of PID controllers

Tuning PID Controller Institute of Industrial Control,

INDUSTRIAL AUTOMATION (Getting Started week -1). Contents PID Controller. Implementation of PID Controller. Response under actuator Saturation. PID with.

CHE 185 – PROCESS CONTROL AND DYNAMICS

Dynamic Energy Balance. Last time: well-mixed CSTR w/flow & reaction for multiple reactions: rxn #

Exercise 3. Solutions From Fujio Kida, JGC Co. All cases

Ch 2.3: Modeling with First Order Equations Mathematical models characterize physical systems, often using differential equations. Model Construction:

1 Dynamic Behavior Chapter 5 In analyzing process dynamic and process control systems, it is important to know how the process responds to changes in the.

Process Control: Designing Process and Control Systems for Dynamic Performance Chapter 18. Level Control Copyright © Thomas Marlin 2013 The copyright holder.

DYNAMIC BEHAVIOR OF PROCESSES :

Background 1. Energy conservation equation If there is no friction.

Probably© the smoothest PID tuning rules in the world: Lower limit on controller gain for acceptable disturbance rejection Sigurd Skogestad Department.

CHAPTER V CONTROL SYSTEMS

Presentation at NI Day April 2010 Lillestrøm, Norway

Control engineering and signal processing

PID-tuning using the SIMC rules

Model-based Predictive Control (MPC)

Mathematical Models for Simulation, Control and Testing

Introduction to process control

Course PEF3006 Process Control Fall 2017 Lecture: Process dynamics

Course IIA1117 Control Engineering Fall 2017 Plant-wide control

State estimation (Kalman filter)

Model-based Predictive Control (MPC)

Tuning of PID controllers

Dynamic Behavior Chapter 5

Av Finn Aakre Haugen IA3112 Automatiseringsteknikk og EK3114 Automatisering og vannkraftregulering Høstsemesteret 2017 Gain scheduling.

Basic Design of PID Controller

Course PEF3006 Process Control Fall 2018 Feedforward Control

Tuning of PID controllers

Servo Tuning for Path Applications

PID controller: Reverse or direct action?

Model-based Predictive Control (MPC)

Model-based Predictive Control (MPC)

Course PEF3006 Process Control Fall 2018 Lecture: Process dynamics

Course PEF3006 Process Control Fall 2018 PID Control

State estimation (Kalman filter)

4.1 Stability Routh Tabulation Example 4.1

Introduction to process control

Pipe Networks Pipeline systems You are here Transmission lines

بسم الله الرحمن الرحيم PID Controllers

IA3112 Automatiseringsteknikk Høsten 2017

Course IIA1117 Control Engineering Fall 2017 Cascade control

Averaging (or equalizing) level control

Y The graph of a rule connecting x and y is shown. From the graph, when x is -1, what is y? Give me a value of x that makes y positive, negative, equal.

Course IIA1117 Control Engineering Fall 2018 Cascade control

Course IIA1117 Control Engineering Fall 2018 Plant-wide control

Av Finn Aakre Haugen IA3112 Automatiseringsteknikk og EK3114 Automatisering og vannkraftregulering Høstsemesteret 2018 Gain scheduling.

Course PEF3006 Process Control Fall 2017 Cascade control

Course PEF3006 Process Control Fall 2017 Feedforward Control

7. Stability [Ref:Rao] 7.1 System stability

How a control system may become unstable

Averaging (or equalizing) level control

Course IIA1117 Control Engineering Fall 2018 Ratio control

Course IIA1117 Control Engineering Fall 2017 Ratio control

On-Off Control (alternative to PID)

Proportional Control and Disturbance Changes

Averaging (or equalizing) level control

Applications of Bernoulli Equations

Course PEF3006 Process Control Fall 2018 Cascade control

Closed-Loop Frequency Response and Sensitivity Functions

On-Off Control (alternative to PID)

How a control system may become unstable

Pipe Networks Pipeline systems You are here Transmission lines

Controller Tuning Relations

Performance and Robustness of the Smith Predictor Controller

PID controller: Reverse or direct action?

Course PEF3006 Process Control Fall 2017 PID Control

Some critical questions

Dynamic Behavior Chapter 5

Presentation transcript:

Av Finn Aakre Haugen (finn.haugen@usn.no) IA3112 Automatiseringsteknikk Høsten 2018 Midlende nivåregulering av buffertank (eng.: averaging level control...) Av Finn Aakre Haugen (finn.haugen@usn.no) USN. IA3112 Auttek. Haugen. 2018.

Level controller (must be tuned for 2 Example of application (results shown at end of this PPT): Level control of equalization magazine upstreams the VEAS water resource recovery facility (wrrf) or resource recovery facility (wrrf), at Slemmestad, south of Oslo, Norway: Inflow Outflow is smoother than inflow! Level controller (must be tuned for "soft (compliant" control) USN. IA3112 Auttek. Haugen. 2018.

Principal buffer tank with a level control system 3 Principal buffer tank with a level control system Simulator USN. IA3112 Auttek. Haugen. 2018.

4 How to tune LC? We need a sluggish or soft or compliant LC so that the liquid volume (the level) can take up the inflow variations. Ziegler-Nichols is useless here since it gives fast or stiff control :-( But Skogestad is excellent, using Tc as tuning parameter :-) Kc = 1/(Ki*Tc) Ti = 2*Tc where Ki = -1/A is the integrator gain or normalized process step response. How to select Tc? . . . USN. IA3112 Auttek. Haugen. 2018.

Δhmax <= (Tc/A)*ΔFin Cont.: How to select Tc? 5 As a start, assume P (proportional) level controller. It can be shown, from a mathematical model of the level control system, that Δhmax = (Tc/A)*ΔFin where Δhmax is corresponding maximum allowed level change (in steady state) after max inflow step change, ΔFin. With a PI controller with the same Tc: Δhmax <= (Tc/A)*ΔFin Solving this inequality for Tc gives Tc >= A* Δhmax/ΔFin => Specification of Tc in the PI settings: Tc = A* Δhmax/ΔFin USN. IA3112 Auttek. Haugen. 2018.

Comparison of responses in level h 6 Comparison of responses in level h due to step change in inflow Fin, with P and with PI controllers: USN. IA3112 Auttek. Haugen. 2018.

Tc = A* Δhmax/ΔFin = 2000*(-0.5)/(-1) = 1000 s 7 Example Assumptions: A = 2000 m2 ΔFin = 1 m3/s Δhmax = 0.5 m Resulting Tc: Tc = A* Δhmax/ΔFin = 2000*(-0.5)/(-1) = 1000 s PI settings: Kc = 1/(Ki*Tc) = - A/Tc = - ΔFin/Δhmax = - (1 m3/s)/0.5 m = - 2.0 (m3/s)/m Ti = 2*Tc = 2*1000 s = 2000 s USN. IA3112 Auttek. Haugen. 2018.

8 Simulator USN. IA3112 Auttek. Haugen. 2018.

Results from VEAS: Pump flow Much smoother pump flow 9 Results from VEAS: With original PI settings in the LC (Kc = 8.0 , Ti = 1000 s) With new (Skogestad) PI settings in the LC (Kc = 3.1 , Ti = 3240 s) Pump flow Much smoother pump flow USN. IA3112 Auttek. Haugen. 2018.