M. Sc. Tim-Lukas Habich

I'm doing research in the area of controlling soft snake robots, which are suitable for medical applications such as endoscopy or biopsy. The (conventional) model-based control of such nonlinear systems is challenging. Despite very high modeling efforts, poor control performance may be present. For this reason, my work focuses on learning-based control approaches based on machine learning methods. Gaussian processes or (recurrent) neural networks are trained using real experimental data and therefore represent the dynamics of arbitrary nonlinear systems. The realization of model predictive controllers using such data-based approaches is the core of my research.