ILC

Iterative learning control

Information

In industrial practice it frequently happens that particular disturbances on control systems occur periodically. These problems may come from actuators with a periodic disturbance torque profile or they can also be due to rollers which are out-of-round and can introduce a periodic disturbance equal to the rotational frequency or to one of its higher harmonics. Moreover in practice a system often executes the same task repeatedly. For instance, if robotic systems have to perform pick-and-place tasks or batch processes where the same temperature profile for each batch should be followed. For such situations 'Iterative learning control' techniques are developed during the past decade. ILC has been applied successfully in many industrial systems resulting in e.g. a high degree of noise suppression.

 

The course provides knowledge and understanding of:

• The operation of control systems specifically designed to perform periodic tasks/set points and for suppressing periodic disturbances.
• The design of learning controllers in which learning rate and stability are weighted one against another.
• The possible applications of these controllers in mechanical positioning systems and in process control.
• It also gives skills in computer aided design of learning controllers with real-time implementation and performance analysis on an industrial mechanical positioning system.

Intended for

This course is for engineers involved in control systems who want to gain more insight into the possibilities and limitations of learning control in an industrial setting. Participants already have a Bachelor or Master education in electrical engineering, mechanical engineering, mechatronics, physics or equivalent practical experience and must have some basic understanding of servo control. This course is particularly intended for engineers having followed the course in 'Motion control tuning'.

Programme

The following topics are treated:

• Overview application areas.
• 'Iterative learning control': principle, convergence, design.
• 'Fixed structure learning feedforward': principle, design, analysis.
• Practical examples: production of glass, hard disk drives, wafer stepper.
• Simulation and design of systems using Matlab and SIMULINK.
• Real-time implementation using SIMULINK and TUe DAC/ QAD instrumentation on HP printer setup.

 

This course is part of the Certified Precision Engineer program of the Dutch Society for Precision Engineering (DSPE).

Timetable

To be determined