This National Science Foundation (NSF) Project Grant under the Computer and Information Science and Engineering (CFDA 47.070) program provides $600,000 to support research at the University of Michigan to develop a comprehensive framework for end-to-end verification of control systems. The key products and services being delivered include:
The project will bridge the gap between high-level control theory and low-level implementation details, providing formal guarantees on the implementation rather than just the high-level algorithm. The research will focus on linear discrete-time systems, hybrid systems with continuous-time plants and discrete-time controllers, and control algorithms using optimization routines like model predictive control. Throughout, the project will emphasize numerical accuracy using machine arithmetic. One major outcome will be the provision of modular, reusable, open-source formal proofs of end-to-end correctness for common controllers like PID and MPC. The techniques will be applied to three domains: car collision avoidance, aircraft collision avoidance, and spacecraft docking. The project also aims to provide research opportunities for underrepresented students and incorporate the findings into control theory and embedded systems coursework.
Generated 3/4/25, 6:50 AM