The course times are not decided yet. 1. Mathematical modeling of uncertainty in linear and nonlinear dynamic systems
2. Stochastic modeling approaches
• Probability distributions
• Bayesian state estimation for discrete-time systems (linear/nonlinear) and for continuous-time systems (linear)
• Linear estimation techniques in an extended state-space (Carleman linearization for special system classes)
• Monte-Carlo methods
3. Estimation of states, parameters and simulation of uncertain processes
• Outlook: Markov models
• Outlook: Bayesian networks
4. Set-based approaches
• Set-based algorithms: Forward-backward contractor and bisection techniques
• Interval methods for a verified solution of ordinary differential equations and for a stability proof of uncertain systems
• Estimation of states and parameters as well as simulation of uncertain processes
5. Outlook: Synthesis of controllers and state observers under an explicit description of uncertainty
The course times are not decided yet. 1. Mathematical modeling of uncertainty in linear and nonlinear dynamic systems
2. Stochastic modeling approaches
• Probability distributions
• Bayesian state estimation for discrete-time systems (linear/nonlinear) and for continuous-time systems (linear)
• Linear estimation techniques in an extended state-space (Carleman linearization for special system classes)
• Monte-Carlo methods
3. Estimation of states, parameters and simulation of uncertain processes
• Outlook: Markov models
• Outlook: Bayesian networks
4. Set-based approaches
• Set-based algorithms: Forward-backward contractor and bisection techniques
• Interval methods for a verified solution of ordinary differential equations and for a stability proof of uncertain systems
• Estimation of states and parameters as well as simulation of uncertain processes
5. Outlook: Synthesis of controllers and state observers under an explicit description of uncertainty
The course times are not decided yet. 1. Mathematical modeling of uncertainty in linear and nonlinear dynamic systems
2. Stochastic modeling approaches
• Probability distributions
• Bayesian state estimation for discrete-time systems (linear/nonlinear) and for continuous-time systems (linear)
• Linear estimation techniques in an extended state-space (Carleman linearization for special system classes)
• Monte-Carlo methods
3. Estimation of states, parameters and simulation of uncertain processes
• Outlook: Markov models
• Outlook: Bayesian networks
4. Set-based approaches
• Set-based algorithms: Forward-backward contractor and bisection techniques
• Interval methods for a verified solution of ordinary differential equations and for a stability proof of uncertain systems
• Estimation of states and parameters as well as simulation of uncertain processes
5. Outlook: Synthesis of controllers and state observers under an explicit description of uncertainty
Hinweise zum Modul
Prüfungszeiten
Am Ende der Veranstaltungszeit
Module examination
Portfolio
Skills to be acquired in this module
Competencies: The students identify fundamentals of uncertainty modeling in control systems as well as problem-specific methods for the consideration of uncertainty during simulation and observer synthesis.
Professional Competences:
The students identify fundamentals of uncertainty modeling in control systems, characterize problem-specific solution techniques for systems with stochastic and set-based uncertainty, and are aware of software implementations in simulation, control, and state estimation.
Methological competences:
The students analyze problems of control-oriented uncertainty modeling, analyze fundamental solution techniques on a theoretical basis as well as transfer and generalize them independently toward novel research-oriented application scenarios.
Social competences:
The students analyze problems of control-oriented uncertainty modeling, analyze fundamental solution techniques on a theoretical basis as well as transfer and generalize them independently toward novel research-oriented application scenarios.
Self competences:
The students critically reflect the achieved results of their project work and acknowledge limitations of various approaches for a control-oriented uncertainty modeling.
