# Search result: Catalogue data in Spring Semester 2021

Robotics, Systems and Control Master | ||||||

Core Courses | ||||||

Number | Title | Type | ECTS | Hours | Lecturers | |
---|---|---|---|---|---|---|

227-0528-00L | Power System Dynamics, Control and Operation | W | 6 credits | 4G | G. Hug | |

Abstract | The electric power system is a system that is never in steady state due to constant changes in load and generation inputs. This course is dedicated to the dynamical properties of the electric power grid including how the system state is estimated, generation/load balance is ensured by frequency control and how the system reacts in case of faults in the system. The course includes two excursions. | |||||

Objective | The learning objectives of the course are to understand and be able to apply the dynamic modeling of power systems, to compute and discuss the actions of generators based on frequency control, to describe the workings of a synchronous machine and the implications on the grid, to describe and apply state estimation procedures, to discuss the IT infrastructure and protection algorithms in power systems. | |||||

Content | The electric power system is a system that is never in steady state due to constant changes in load and generation inputs. Consequently, the monitoring and operation of the electric power grid is a challenging task. The course starts with the introduction of general operational procedures and the discussion of state estimation which is an important tool to observe the state of the grid. The course is then dedicated to the modeling and studying of the dynamical properties of the electric power grid. Frequency control which ensures the generation/load balance in real time is the basis for real-time control and is presented in depth. For the analysis of how the system detects and reacts dynamically in fault situations, protection and dynamic models for synchronous machines are introduced. | |||||

Lecture notes | Lecture notes. WWW pages. | |||||

227-0560-00L | Deep Learning for Autonomous Driving Registration in this class requires the permission of the instructors. Class size will be limited to 80 students. Please send an email to Dengxin Dai <dai@vision.ee.ethz.ch> about your courses/projects that are related to machine learning, computer vision, and Robotics. | W | 6 credits | 3V + 2P | D. Dai, A. Liniger | |

Abstract | Autonomous driving has moved from the realm of science fiction to a very real possibility during the past twenty years, largely due to rapid developments of deep learning approaches, automotive sensors, and microprocessor capacity. This course covers the core techniques required for building a self-driving car, especially the practical use of deep learning through this theme. | |||||

Objective | Students will learn about the fundamental aspects of a self-driving car. They will also learn to use modern automotive sensors and HD navigational maps, and to implement, train and debug their own deep neural networks in order to gain a deep understanding of cutting-edge research in autonomous driving tasks, including perception, localization and control. After attending this course, students will: 1) understand the core technologies of building a self-driving car; 2) have a good overview over the current state of the art in self-driving cars; 3) be able to critically analyze and evaluate current research in this area; 4) be able to implement basic systems for multiple autonomous driving tasks. | |||||

Content | We will focus on teaching the following topics centered on autonomous driving: deep learning, automotive sensors, multimodal driving datasets, road scene perception, ego-vehicle localization, path planning, and control. The course covers the following main areas: I) Foundation a) Fundamentals of a self-driving car b) Fundamentals of deep-learning II) Perception a) Semantic segmentation and lane detection b) Depth estimation with images and sparse LiDAR data c) 3D object detection with images and LiDAR data d) Object tracking and Lane Detection III) Localization a) GPS-based and Vision-based Localization b) Visual Odometry and Lidar Odometry IV) Path Planning and Control a) Path planning for autonomous driving b) Motion planning and vehicle control c) Imitation learning and reinforcement learning for self driving cars The exercise projects will involve training complex neural networks and applying them on real-world, multimodal driving datasets. In particular, students should be able to develop systems that deal with the following problems: - Sensor calibration and synchronization to obtain multimodal driving data; - Semantic segmentation and depth estimation with deep neural networks ; - 3D object detection and tracking in LiDAR point clouds | |||||

Lecture notes | The lecture slides will be provided as a PDF. | |||||

Prerequisites / Notice | This is an advanced grad-level course. Students must have taken courses on machine learning and computer vision or have acquired equivalent knowledge. Students are expected to have a solid mathematical foundation, in particular in linear algebra, multivariate calculus, and probability. All practical exercises will require basic knowledge of Python and will use libraries such as PyTorch, scikit-learn and scikit-image. | |||||

227-0694-00L | Game Theory and Control | W | 4 credits | 2V + 2U | S. Bolognani | |

Abstract | Game Theory is the study of strategic decision making, and was used to solve problems in economics by John Nash (A Beautiful Mind) and others. We study concepts and methods in Game Theory, and show how these can be used to solve control design problems. The course covers non-cooperative dynamic games and Nash equilibria, and emphasizes their use in control applications. | |||||

Objective | Formulate an optimal control problem as a noncooperative dynamic game, compute mixed and behavioural strategies for different equilibria. | |||||

Content | Introduction to game theory, mathematical tools including convex optimisation and dynamic programming, zero sum games in matrix and extensive form, pure and mixed strategies, minimax theorem, nonzero sum games in normal and extensive form, numerical computation of mixed equilibrium strategies, Nash and Stackelberg equilibria, potential games, infinite dynamic games, differential games, behavioral strategies and informational properties for dynamic games, aggregative games, VCG mechanism. | |||||

Lecture notes | Will be made available from SPOD or course webpage. | |||||

Literature | Basar, T. and Olsder, G. Dynamic Noncooperative Game Theory, 2nd Edition, Society for Industrial and Applied Mathematics, 1998. Available through ETH Bibliothek directly at http://epubs.siam.org/doi/abs/10.1137/1.9781611971132. | |||||

Prerequisites / Notice | Control Systems I (or equivalent). Necessary methods and concepts from optimization will be covered in the course. | |||||

227-0696-00L | Predictive Control of Power Electronics Systems | W | 6 credits | 2V + 2U | T. Geyer | |

Abstract | Bridging the gap between modern control methods and power electronics, this course focuses on predictive control methods applied to power electronics systems. This includes model predictive control methods with and without a modulator. This course targets power electronics and control students. | |||||

Objective | - Knowledge of modern time-domain control methods applied to three-phase converters and their corresponding loads. These control methods include model predictive control (MPC) and deadbeat control. - Understanding of optimized pulse patterns and techniques to achieve fast closed-loop control. - Ability to derive suitable mathematical models. - Knowledge of and experience in optimization techniques to solve the underlying mixed-integer and quadratic programs. - Appreciation of the advantages and disadvantages of the different control methods. | |||||

Content | - Review of mathematical modelling and time-domain control methods (particularly MPC and deadbeat control). - Direct MPC with reference tracking (finite control set MPC). Derivation of mathematical models of three-phase power electronics systems, formulation of the control problem, techniques to solve the one-step and the multi-step horizon problems using branch and bound techniques. - MPC with optimized pulse patterns (OPPs). Computation of OPPs, formulation of fast closed-loop controllers and methods to solve the underlying quadratic programming problem. - Indirect MPC with pulse width modulation (PWM). Formulation of the MPC problem, imposition of hard and soft constraints, techniques to solve the quadratic program in real time and application to modular multilevel converters. - Summary of recent research results and activities. - Matlab / Simulink exercises to enhance the understanding of the control concepts. | |||||

Lecture notes | The lecture is based on the book "Model Predictive Control of High Power Converters and Industrial Drives" by T. Geyer. Additional notes will be made available in the class. | |||||

Prerequisites / Notice | - Power Electronic Systems I - Control Systems I (Regelsysteme I) - Signal and System Theory II | |||||

252-0220-00L | Introduction to Machine Learning Limited number of participants. Preference is given to students in programmes in which the course is being offered. All other students will be waitlisted. Please do not contact Prof. Krause for any questions in this regard. If necessary, please contact studiensekretariat@inf.ethz.ch | W | 8 credits | 4V + 2U + 1A | A. Krause, F. Yang | |

Abstract | The course introduces the foundations of learning and making predictions based on data. | |||||

Objective | The course will introduce the foundations of learning and making predictions from data. We will study basic concepts such as trading goodness of fit and model complexitiy. We will discuss important machine learning algorithms used in practice, and provide hands-on experience in a course project. | |||||

Content | - Linear regression (overfitting, cross-validation/bootstrap, model selection, regularization, [stochastic] gradient descent) - Linear classification: Logistic regression (feature selection, sparsity, multi-class) - Kernels and the kernel trick (Properties of kernels; applications to linear and logistic regression); k-nearest neighbor - Neural networks (backpropagation, regularization, convolutional neural networks) - Unsupervised learning (k-means, PCA, neural network autoencoders) - The statistical perspective (regularization as prior; loss as likelihood; learning as MAP inference) - Statistical decision theory (decision making based on statistical models and utility functions) - Discriminative vs. generative modeling (benefits and challenges in modeling joint vy. conditional distributions) - Bayes' classifiers (Naive Bayes, Gaussian Bayes; MLE) - Bayesian approaches to unsupervised learning (Gaussian mixtures, EM) | |||||

Literature | Textbook: Kevin Murphy, Machine Learning: A Probabilistic Perspective, MIT Press | |||||

Prerequisites / Notice | Designed to provide a basis for following courses: - Advanced Machine Learning - Deep Learning - Probabilistic Artificial Intelligence - Seminar "Advanced Topics in Machine Learning" | |||||

252-0312-00L | Ubiquitous Computing | W | 6 credits | 2V + 3A | C. Holz | |

Abstract | Ubiquitous Computing means interacting with information and with each other anywhere, mediated through miniature technology everywhere. We will investigate the technical aspects of Ubicomp, particularly sensing, processing, and sense making: input (touch & gesture), activity, monitoring cardiovascular health and neurological conditions, context & location sensing, affective computing. | |||||

Objective | The course will combine high-level concepts with low-level technical methods needed to sense, detect, and understand them. High-level: – input modalities for interactive systems (touch, gesture) – "activities" and "events" (exercises and other mechanical activities such as movements and resulting vibrations) – health monitoring (basic cardiovascular physiology) – location (GPS, urban simulations, smart cities and development) – affective computing (emotions, mood, personality) Low-level: – sampling (Shannon Nyquist) and filtering (FIR, IIR), time and frequency domains (Fourier transforms) – cross-modal sensor systems, signal synchronization and correlation – event detection, classification, prediction using basic signal processing as well as learning-based methods – sensor types: optical, mechanical/acoustic, electromagnetic – signals modalities and processing of: application (modalities/methods) * touch detection (resistive sensing, capacitive sensing, diffuse illumination/DI, spectral reflections, frustrated total internal reflection/FTIR, fingerprint scanning, surface-acoustic waves) * gesture recognition (inertial sensing through accelerometers, gyroscopes) * activity detection and tracking (inertial, acoustic, vibrotactile for classification, counting, vibrometry) * occupation and use (electricity monitoring, water consumption, single-point sensing) * cardiovascular (electrocardioagraphy, photoplethysmography, pulse oximetry, ballistocardiography, blood pressure, pulse transit time, bio impedance) * affective computing (heart rate variability, R-R intervals, electrodermal activity, sympathetic tone, facial expressions) * neurological (fatigue, fatigability) * location (GPS, BLE, Wifi) | |||||

Content | "The most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it" — Mark Weiser, 1991. This is the premise of Ubiquitous Computing, a vision that is slowly becoming reality as everything is a device and we can interact with information and with each other anywhere, mediated through miniature technology. Along with this change, interaction modalities have changed, too, from explicit input on keyboards and mice to implicit and passively observed input through sensors in the environment (e.g., speakers, cameras, temperature/occupancy detectors) and those we now wear on our bodies (e.g., health sensors, activity sensors, miniature computers we call smartwatches). In this course, we will look at the technical side of Ubicomp, particularly – sensing (incl. 'signals', sampling, data acquisition methods, controlled user studies, uncontrolled studies in-the-wild), – processing (incl. frequencies, feature extraction, detection), and – sense making: input sensing (touch & gesture), activity sensing (motion), monitoring cardiovascular health, affective state, neurological conditions (with basics on cardiovascular physiology + PPG, PulseOx, ECG, EDA, BCG, SCG, HRV, BioZ, IPG, PAT, PTT), context & location sensing (GPS/Wifi, motion). Lectures will be accompanied by practical sessions that focus on sensor modalities and signal processing. Here, we will work on existing data sets and devise methods to record our own data for processing and prediction purposes. A series of reading assignments, covering both well-established publications in Ubicomp as well as emerging results and methods, will bridge the fundamentals and topics taught in class to academic research and real-world problems. More information on the course site: https://teaching.siplab.org/ubiquitous_computing/2021/ | |||||

Lecture notes | Copies of slides will be made available. Lectures will be recorded and made available online. More information on the course site: https://teaching.siplab.org/ubiquitous_computing/2021/ | |||||

Literature | Will be provided in the lecture. To put you in the mood: Mark Weiser: The Computer for the 21st Century. Scientific American, September 1991, pp. 94-104 | |||||

252-0526-00L | Statistical Learning Theory | W | 8 credits | 3V + 2U + 2A | J. M. Buhmann, C. Cotrini Jimenez | |

Abstract | The course covers advanced methods of statistical learning: - Variational methods and optimization. - Deterministic annealing. - Clustering for diverse types of data. - Model validation by information theory. | |||||

Objective | The course surveys recent methods of statistical learning. The fundamentals of machine learning, as presented in the courses "Introduction to Machine Learning" and "Advanced Machine Learning", are expanded from the perspective of statistical learning. | |||||

Content | - Variational methods and optimization. We consider optimization approaches for problems where the optimizer is a probability distribution. We will discuss concepts like maximum entropy, information bottleneck, and deterministic annealing. - Clustering. This is the problem of sorting data into groups without using training samples. We discuss alternative notions of "similarity" between data points and adequate optimization procedures. - Model selection and validation. This refers to the question of how complex the chosen model should be. In particular, we present an information theoretic approach for model validation. - Statistical physics models. We discuss approaches for approximately optimizing large systems, which originate in statistical physics (free energy minimization applied to spin glasses and other models). We also study sampling methods based on these models. | |||||

Lecture notes | A draft of a script will be provided. Lecture slides will be made available. | |||||

Literature | Hastie, Tibshirani, Friedman: The Elements of Statistical Learning, Springer, 2001. L. Devroye, L. Gyorfi, and G. Lugosi: A probabilistic theory of pattern recognition. Springer, New York, 1996 | |||||

Prerequisites / Notice | Knowledge of machine learning (introduction to machine learning and/or advanced machine learning) Basic knowledge of statistics. | |||||

252-0579-00L | 3D Vision | W | 5 credits | 3G + 1A | M. Pollefeys, V. Larsson | |

Abstract | The course covers camera models and calibration, feature tracking and matching, camera motion estimation via simultaneous localization and mapping (SLAM) and visual odometry (VO), epipolar and mult-view geometry, structure-from-motion, (multi-view) stereo, augmented reality, and image-based (re-)localization. | |||||

Objective | After attending this course, students will: 1. understand the core concepts for recovering 3D shape of objects and scenes from images and video. 2. be able to implement basic systems for vision-based robotics and simple virtual/augmented reality applications. 3. have a good overview over the current state-of-the art in 3D vision. 4. be able to critically analyze and asses current research in this area. | |||||

Content | The goal of this course is to teach the core techniques required for robotic and augmented reality applications: How to determine the motion of a camera and how to estimate the absolute position and orientation of a camera in the real world. This course will introduce the basic concepts of 3D Vision in the form of short lectures, followed by student presentations discussing the current state-of-the-art. The main focus of this course are student projects on 3D Vision topics, with an emphasis on robotic vision and virtual and augmented reality applications. | |||||

263-5806-00L | Computational Models of Motion | W | 8 credits | 2V + 2U + 3A | S. Coros, M. Bächer, B. Thomaszewski | |

Abstract | This course covers fundamentals of physics-based modelling and numerical optimization from the perspective of character animation and robotics applications. The methods discussed in class derive their theoretical underpinnings from applied mathematics, control theory and computational mechanics, and they will be richly illustrated using examples ranging from locomotion controllers and crowd simula | |||||

Objective | Students will learn how to represent, model and algorithmically control the behavior of animated characters and real-life robots. The lectures are accompanied by programming assignments (written in C++) and a capstone project. | |||||

Content | Optimal control and trajectory optimization; multibody systems; kinematics; forward and inverse dynamics; constrained and unconstrained numerical optimization; mass-spring models for crowd simulation; FEM; compliant systems; sim-to-real; robotic manipulation of elastically-deforming objects. | |||||

Prerequisites / Notice | Experience with C++ programming, numerical linear algebra and multivariate calculus. Some background in physics-based modeling, kinematics and dynamics is helpful, but not necessary. | |||||

376-1217-00L | Rehabilitation Engineering I: Motor Functions | W | 4 credits | 2V + 1U | S. Raspopovic, M. Xiloyannis | |

Abstract | “Rehabilitation engineering” is the application of science and technology to ameliorate the handicaps of individuals with disabilities in order to reintegrate them into society. The goal of this lecture is to present classical and new rehabilitation engineering principles and examples applied to compensate or enhance especially motor deficits. | |||||

Objective | Provide theoretical and practical knowledge of principles and applications used to rehabilitate individuals with motor disabilities. | |||||

Content | “Rehabilitation” is the (re)integration of an individual with a disability into society. Rehabilitation engineering is “the application of science and technology to ameliorate the handicaps of individuals with disability”. Such handicaps can be classified into motor, sensor, and cognitive (also communicational) disabilities. In general, one can distinguish orthotic and prosthetic methods to overcome these disabilities. Orthoses support existing but affected body functions (e.g., glasses, crutches), while prostheses compensate for lost body functions (e.g., cochlea implant, artificial limbs). In case of sensory disorders, the lost function can also be substituted by other modalities (e.g. tactile Braille display for vision impaired persons). The goal of this lecture is to present classical and new technical principles as well as specific examples applied to compensate or enhance mainly motor deficits. Modern methods rely more and more on the application of multi-modal and interactive techniques. Multi-modal means that visual, acoustical, tactile, and kinaesthetic sensor channels are exploited by displaying the patient with a maximum amount of information in order to compensate his/her impairment. Interaction means that the exchange of information and energy occurs bi-directionally between the rehabilitation device and the human being. Thus, the device cooperates with the patient rather than imposing an inflexible strategy (e.g., movement) upon the patient. Multi-modality and interactivity have the potential to increase the therapeutical outcome compared to classical rehabilitation strategies. In the 1 h exercise the students will learn how to solve representative problems with computational methods applied to exoprosthetics, wheelchair dynamics, rehabilitation robotics and neuroprosthetics. | |||||

Literature | Introductory Books Neural prostheses - replacing motor function after desease or disability. Eds.: R. Stein, H. Peckham, D. Popovic. New York and Oxford: Oxford University Press. Advances in Rehabilitation Robotics – Human-Friendly Technologies on Movement Assistance and Restoration for People with Disabilities. Eds: Z.Z. Bien, D. Stefanov (Lecture Notes in Control and Information Science, No. 306). Springer Verlag Berlin 2004. Intelligent Systems and Technologies in Rehabilitation Engineering. Eds: H.N.L. Teodorescu, L.C. Jain (International Series on Computational Intelligence). CRC Press Boca Raton, 2001. Control of Movement for the Physically Disabled. Eds.: D. Popovic, T. Sinkjaer. Springer Verlag London, 2000. Interaktive und autonome Systeme der Medizintechnik - Funktionswiederherstellung und Organersatz. Herausgeber: J. Werner, Oldenbourg Wissenschaftsverlag 2005. Biomechanics and Neural Control of Posture and Movement. Eds.: J.M. Winters, P.E. Crago. Springer New York, 2000. Selected Journal Articles Abbas, J., Riener, R. (2001) Using mathematical models and advanced control systems techniques to enhance neuroprosthesis function. Neuromodulation 4, pp. 187-195. Burdea, G., Popescu, V., Hentz, V., and Colbert, K. (2000): Virtual reality-based orthopedic telerehabilitation, IEEE Trans. Rehab. Eng., 8, pp. 430-432 Colombo, G., Jörg, M., Schreier, R., Dietz, V. (2000) Treadmill training of paraplegic patients using a robotic orthosis. Journal of Rehabilitation Research and Development, vol. 37, pp. 693-700. Colombo, G., Jörg, M., Jezernik, S. (2002) Automatisiertes Lokomotionstraining auf dem Laufband. Automatisierungstechnik at, vol. 50, pp. 287-295. Cooper, R. (1993) Stability of a wheelchair controlled by a human. IEEE Transactions on Rehabilitation Engineering 1, pp. 193-206. Krebs, H.I., Hogan, N., Aisen, M.L., Volpe, B.T. (1998): Robot-aided neurorehabilitation, IEEE Trans. Rehab. Eng., 6, pp. 75-87 Leifer, L. (1981): Rehabilitive robotics, Robot Age, pp. 4-11 Platz, T. (2003): Evidenzbasierte Armrehabilitation: Eine systematische Literaturübersicht, Nervenarzt, 74, pp. 841-849 Quintern, J. (1998) Application of functional electrical stimulation in paraplegic patients. NeuroRehabilitation 10, pp. 205-250. Riener, R., Nef, T., Colombo, G. (2005) Robot-aided neurorehabilitation for the upper extremities. Medical & Biological Engineering & Computing 43(1), pp. 2-10. Riener, R., Fuhr, T., Schneider, J. (2002) On the complexity of biomechanical models used for neuroprosthesis development. International Journal of Mechanics in Medicine and Biology 2, pp. 389-404. Riener, R. (1999) Model-based development of neuroprostheses for paraplegic patients. Royal Philosophical Transactions: Biological Sciences 354, pp. 877-894. | |||||

Prerequisites / Notice | Target Group: Students of higher semesters and PhD students of - D-MAVT, D-ITET, D-INFK - Biomedical Engineering - Medical Faculty, University of Zurich Students of other departments, faculties, courses are also welcome | |||||

227-0690-12L | Advanced Topics in Control (Spring 2021)New topics are introduced every year. | W | 4 credits | 2V + 2U | F. Dörfler, M. Hudoba de Badyn, W. Mei | |

Abstract | Advanced Topics in Control (ATIC) covers advanced research topics in control theory. It is offered each Spring semester with the topic rotating from year to year. Repetition for credit is possible, with consent of the instructor. During the spring of 2020, the course will cover a range of topics in distributed systems control. | |||||

Objective | By the end of this course you will have developed a sound and versatile toolkit to tackle a range of problems in network systems and distributed systems control. In particular, we will develop the methodological foundations of algebraic graph theory, consensus algorithms, and multi-agent systems. Building on top of these foundations we cover a range of problems in epidemic spreading over networks, swarm robotics, sensor networks, opinion dynamics, distributed optimization, and electrical network theory. | |||||

Content | Distributed control systems include large-scale physical systems, engineered multi-agent systems, as well as their interconnection in cyber-physical systems. Representative examples are electric power grids, swarm robotics, sensor networks, and epidemic spreading over networks. The challenges associated with these systems arise due to their coupled, distributed, and large-scale nature, and due to limited sensing, communication, computing, and control capabilities. This course covers algebraic graph theory, consensus algorithms, stability of network systems, distributed optimization, and applications in various domains. | |||||

Lecture notes | A complete set of lecture notes and slides will be provided. | |||||

Literature | The course will be largely based on the following set of lecture notes co-authored by one of the instructors: http://motion.me.ucsb.edu/book-lns/ | |||||

Prerequisites / Notice | Sufficient mathematical maturity, in particular in linear algebra and dynamical systems. |

- Page 2 of 2 All