Towards Real-time Adaptive Approximation
Abstract
Cyber-physical systems (CPS) are real-time systems that operate in dynamic and non-deterministic environments. Models are often used for control and prediction, however do not reason on the trade-off between real-time constraints and uncertainty. This paper presents a conceptual model to reason on adaptive approximation in such systems. Furthermore, we envision a framework to allow the adaptivity of models, balancing between uncertainty and the real-time behavior of the system.
Origin : Files produced by the author(s)