Learn to analyse nonlinear material behaviour and large deformations using the finite element method. Apply models in Abaqus to assess materials, forming processes and structures.
System Identification with Parameter Estimation and Machine Learning focuses on building mathematical models or models with physical parameters of dynamic systems based on measured input-output data. In this course, you learn how to select a suitable model structure, estimate model parameters, design identification experiments and evaluate whether the resulting model is reliable for practical use.
The course covers theory and application of both classical system identification and newer machine learning-based approaches. You will work with models in time and frequency domain and learn how to estimate and validate models using established methods such as subspace identification and prediction error methods. This makes the course highly relevant for professionals who need models that provide the right level of insight for analysis, control design or system optimisation.
In many engineering applications, reliable models are needed to understand, predict and improve system behaviour. In cases where well-established model equations are available, parameter estimation methods are used to determine the model parameters from experimental data. However, not all systems can be modelled accurately from physical laws alone, especially when the dynamics are complex or partially unknown. System identification addresses this challenge by estimating mathematical models from measured data. Furthermore, as data-driven engineering and machine learning continue to grow in importance, professionals increasingly need to combine classical modelling approaches with modern estimation and learning techniques.
This is a regular master-level course in which both students and professionals can participate. The course combines lectures, tutorials and assignments. Topics include signal and system relations in time and frequency domain, non-parametric identification, parameter estimation in linear-in-the-parameters models, subspace identification, prediction error methods, open-loop versus closed-loop identification, and experiment design.
The course includes one assignments, which may be completed in groups. This assignment is problem-based and can be linked to practical cases, including cases with an industry connection.
This course is intended for professionals who work with (dynamic) systems and need to model, analyse or improve system behaviour based on data. Typical roles include, but are not limited to, control engineers, mechatronics engineers, robotics engineers, system engineers, R&D engineers, data-driven modelling specialists and technical researchers. The course is particularly relevant for professionals in high-tech systems, automation, robotics, process engineering and advanced manufacturing.
Participants should already have a foundation in system analysis or an equivalent background in dynamic systems. More specifically, prior knowledge is required regarding (mathematical) system models such as (ordinary) differential equations, transfer functions, and state-space equations, as well as the relationship between the time and frequency domains. In addition, various concepts from linear algebra are used. The level of the course System Analysis from the Mechanical Engineering BSc program (or an equivalent course) is adequate.
If you successfully complete the course, you will receive a microcredential. This microcredential is a digital record confirming that you have passed the assessment components of the regular master’s course and achieved the corresponding learning outcomes.
You may also participate without completing the assessments. If you choose this option, or do not pass all assessment components, you will receive a digital certificate of participation.
Start date: Mid-April (UT block 2B, Q4)
Total duration: April to June/July (excluding resits) (140 hour studyload)
Number of teaching days / sessions: Lectures + tutorials + 1 assignment
Total programme fee: €2067,15 (5 ECTS *413.43).
Programme type: master course
Location: University of Twente, Enschede, NL
Language of instruction: English
Admission requirements: prior knowledge equivalent to System Analysis (202000128), part of module 5 ME Dynamic Systems (or equivalent). Professionals can receive a list of required knowledge and skills in advance.
Teachers: