Motion control is essential in any application where accurate and fast movements take place, or where motional disturbances should actively be attenuated. Important examples include positioning a product in a manufacturing line, printing on a sheet of paper, lithographic imaging processes, as well as precision scientific instruments such as atomic force microscopes or astronomical instruments. In this course, you will learn how to perfectly tune such a motion control system in a couple of minutes.
To this end, you will learn both the application aspects of mechanical systems, as well as the required theoretical foundations, where Nyquist and Bode diagrams are demystified. The course consists of alternated sessions of theory and application to a motion system, where you will immediately be able to apply and test your newly developed knowledge in practice.
The course content includes the fast identification of frequency response function models in closed-loop, and appreciate their usefulness by comparing these with time domain approaches. You will be able to interpret these frequency response functions and link them to the physical behavior of the mechanical system, where collocated and non-collocated actuators and sensors are a key aspect. The next aspect is to use these frequency response function models for designing controllers. You will learn to tune PID (proportional-integral-derivative) filters, as well as notch and low-pass filters. Frequency domain techniques are used to specify requirements and assessing closed-loop stability, for which you will learn how to use Bode diagrams and Nyquist plots. Your newly developed skills will be directly tested in the loop-shaping game, where you are challenged to tune for the highest performance and compete with the other attendees of the course. In the final part of the course, the performance limitations are further investigated, and further appreciation is given to the collocated and non-collocated control situation. Furthermore, the feedback controller you have designed is complemented with a feedforward controller. Again, a systematic tuning procedure is developed in the same spirit: it allows you to perfectly tune the feedforward controller in a few minutes. In the final session, an outlook is given on extensions, including multivariable feedback loops and learning techniques for automated tuning.
- The training is also available for in-company sessions.