https://sites.google.com/site/chengchihhong/home/teaching/hildesheim_2023_ws_intro_cps (materials)
Description: Cyber-physical systems control and regulate many applications in everyday life, from smart refrigerators, through elevator control, traffic control and safety, automotive systems, etc. to critical and distributed infrastructures. This course provides an understanding of the problems an engineer will face when designing such systems: the selection of the scheduling algorithms, the interfacing with physical sensors and actuators, the modelling of the system and the interaction between its software and hardware components, as well as the analysis of its reachability and real-time performance.
Importantly, the lecture will emphasize the mathematical aspects of system design and analysis, where many topics in the cyber-physical systems such as Kalman-filters, model-predictive control, estimation of the WCET, or even scheduling for synchronous dataflow systems, are based on rigorous mathematical modelling and the application of constraint optimization techniques.
Apart from exercises, the students are expected to apply these knowledge in their own defined projects (e.g., 1/10 autonomous cars), where principles listed in the lecture should be included. Guest lectures from industry speakers will also be planned.
Contents covered in the lecture:
• Model-based Design
• Modelling Continuous Dynamics
• Modeling Discrete Dynamics
• Extended and Hybrid Automata
• Synchronous Reactive (SR) & Static Dataflow (SDF)
• Sensors & Actuators
• Autonomous CPS: Embedded vision with neural networks
• Autonomous CPS: Kalman filter and extensions
• Autonomous CPS: Model-predictive control
• Scheduling and scheduling Anomalies
• Worst-case execution time analysis
• Specification and temporal logic
• Reachability analysis
• Safety assurance for autonomous CPS
Grading (subject to the availability of equipments):
• Exam 40%
• HW 20%
• Project 40%
• Paper presentation from top tier CPS conferences (e.g., ICCPS, EMSOFT) as bonus
Guest speakers in SS2023, TU München
• Mathworks on MBD
• ABB on industry 4.0
• DENSO on scenario-based testing for AD
|
LSF
– Lehre - Studium - Forschung
