Below are a few examples of customized in-house short courses taught to small groups (10-15) of engineers and scientists at customer locations. The customers are usually research and development groups involved in process optimization and control projects that need a group update on the principles of process modeling, dynamics and control through examples specifically chosen among processes they are focused on. From the software point of view SIMULINK and MATLAB from Mathworks are usually the tools of choice for these groups and therefore the courses are also customized to teach software principles and practices as much as needed. Software instructions are always interactive and from an engineering point of view. Courses are  conducted with full participation of everyone at the pace of the group with PCs assigned to teams of two. A typical course day is 30% instruction and 70% exercises and problem solutions on the PCs. Instructor's PC display is projected on the screen or wall for everyone to follow. Class participants get hardcopy and Power Point version of the presentations and software copies of exercises and simulation models. Once the need for a custom class is identified, IETek works with the project lead to determine learning objectives, technical level and exercise examples to propose a class itinerary, which after an iterative review process becomes the starting class agenda. During the progress of instructions some adjustments are usually made on the pace and contents of coverage depending on identified opportunities.


For additional information or questions please contact

IETek

5533 Beverly Ave NE, Tacoma WA 98422-1402, USA

Tel: (253) 925-2179,  Fax: (253) 925-5023

fkayihan@ietek.net


EXAMPLE CLASS AGENDAS:

 

Three day short course on Process Control

Learning objectives:

  • Dynamic behavior
  • Feedback control
  • Model-based control
  • Model predictive control

Day 1:

Process Dynamics and Simulation

  • Process models
  • Model simulation (time domain, Laplace, discrete)
  • Introduction to SIMULINK and MATLAB
  • Noise, sampling and filtering
  • Stochastic models and identification
  • Stability
  • Linearization of nonlinear systems

Single Loop control

  • PID
  • Controller tuning
  • Process reaction curve, Cohen and Coon, Ziegler Nichols, relay

Discrete form

Difficulties and extensions

  • Disturbances / feedforward
  • Time delay / Smith predictor
  • Multivariable
  • Noise filter
  • Constraints

Day 2:

Single loop design with a process model (IMC)

  • Structural analysis
  • PID tuning with IMC
  • Unstable systems
  • RHS zeros
  • Constraints
  • Time delays

Multivariable dynamics and control

  • Interactions
  • Stability
  • Decoupling design
  • Loop pairing
  • Multivariable IMC
  • Limitations

Day 3:

Model predictive control (MPC)

  • Motivation
  • Components
  • Vendors
  • Model for MPC
  • Process Identification
  • Unconstraint MPC
  • Properties
  • Constraint MPC
  • Non-square systems
  • State estimation
  • Nonlinear MPC

 

 

Four day short course on Process Dynamics and Control with Pulp and Paper Applications

Day 1:

  • Introduction to SIMULINK and MATLAB , demonstrative simple engineering problems
  • Linear 1st order system, step and impulse responses, random disturbances, simulation with SIMULINK and MATLAB

Day 2:

  • Material balance for fiber & water mixture systems, dynamics and non-linearity of valves, interpolation and table-lookup in simulation
  • Simulation examples with stock-prep units, first order + dead time model identification
  • Controller tuning (Cohen & Coon, Zeigler-Nichols, Relay), Simulink exercises

Day 3:

  • Smith Predictor
  • Internal Model Control (IMC), Lambda tuning, PI equivalence for 1st order systems
  • High density and blend chest controls (flow, consistency and level), running bump-tests to obtain simple models
  • Multivariable interactions and decoupling, matrix transfer functions, IMC control
  • 2x2 pressure and flow control around a screen (knotter)

Day 4:

  • Dual tank (in series) rate and level control interactions and consistency variability (blend/machine chest example)
  • Data statistics, autocorrelation, confidence limits
  • Basic philosophy and approach of model-based controllers
  • Introduction to paper machine data analysis

IETek 1996-2002, all rights reserved.


 
Home ] Advanced Digester Model ] Advanced Digester Toolbox ] Batch Digester ] Paper Machine Monitoring & Control Workshop ] ACC 2000 Monitoring Workshop ] DOE Digester Control Project ] Digester Benchmark Model v1 ] Digester Benchmark Toolbox v1 ] [ Custom Short Courses ] UDel PCMC ] Publications ] Projects ]