Prognostics and Health Management

What it is about

You will learn methods for forecasting the degradation behavior of technical systems and for estimating their remaining service life. This includes physics-based and data-driven procedures as well as methods for uncertainty estimation. In the end, you can combine models and data to estimate remaining usage durations and prepare suitable maintenance decisions.

What you will be able to do

• Estimate remaining lifetime of systems by combining models and data
• Apply nonlinear least squares methods for parameter estimation
• Perform Bayesian parameter estimation with MCMC methods and particle filters
• Use machine learning methods for data-driven forecasts
• Assess uncertainties in ML-based degradation forecasts
• Assess advantages and disadvantages of physics-based vs. data-driven approaches
• Implement forecasting methods in Matlab

What the module consists of

• Lecture with integrated exercisesConveying the theory and solving exercises; parts of the content are derived at the board, others are presented graphically on slides
• Tutorial exercisesDeepening and application of the methods; solutions are provided in Moodle and discussed in parts during the course

Teaching method

• Blackboard presentationEnables step-by-step derivation of theoretical concepts and formulas for deeper understanding
• Lecture slidesGraphic representation for better illustration of complex content
• Practice tasks with MatlabPractical application and consolidation of the methods; MATLAB code examples support the implementation