Marcus Thelander Andrén
About Me
Ph.D student at the department of Automatic Control since August 2015.
My field of research is event-based control and state estimation. I investigate optimization of event-trigger conditions for event-based estimation and control schemes in a stochastic setting. To sample and actuate optimally based on the current state of the process, rather than based on time, allows for a better performance and lower resource utilization than periodic schemes for the same average sampling rate.
You can find my personal page in Lund University's research portal here.
Research
Project
I am part of the project "Event-Based Control of Stochastic Systems with Application to Server Systems", funded by the Swedish research council.
Code
rbf-approx
Repository with demo code of the numerical method described in Thelander Andrén. "Using Radial Basis Functions to Approximate the LQG-Optimal Sampling Policy." ECC 2019.
Link: rbf-approx
optimal-trigger-bound
Repository with implementation of the numerical method described in Thelander Andrén et al. "On Event-Based Sampling for LQG-Optimal Control." CDC 2017.
Link: optimal-trigger-bound
Systems where the state vector is intermittently reset play a fundamental role in the study of event-based control and estimation. Above is a classic 1D example of an integrator, where the state is reset to the origin whenever it hits the boundaries. The plot shows the propagation of the state's probability density over time. The density starts out Gaussian, but converges to a triangular distribution.
Teaching
HT15
Automatic Control Basic Course
VT16
Automatic Control Basic Course
HT16
Multivariable Control
Projects in Automatic Control
VT17
Mathematical Modelling, Advanced Course
HT17
Automatic Control Basic Course at Beihang University
VT18
Real-Time Systems
HT18
Projects in Automatic Control
VT19
Mathematical Modelling, Advanced Course
Real-Time Systems
HT19
Multivariable Control
