A Smart Adaptable Architecture Based on Contexts for Cyber Physical Systems

The key challenge today isn’t in manufacturing circuits but in programming the massively distributed system that will result from putting all the units together to manage daily huge amount of data with dimensionality factors depending on contexts. The use of the resulting information is even more critical. In the model for describing a particular context property, the domain of interpretation for the property represents the values that it may assume. Hierarchical Formal Concept Analysis (HFCA) models the world of data through the use of contextual objects and attributes (tags) structured in contexts. To evaluate the significance of a concept in a context we compute the significance score and we learn high-dimensional binary feature vectors through the Neural Modeling Fields (NMF) algorithm. The adaptive evolution of context models describes dynamics with different complexity. Each dynamic mode is associated with a mode behavior, the set of trajectories that satisfies the dynamical laws of that mode in a context. A switching signal (an event) determines when a transition occurs between dynamic modes. Symbolic control of nonlinear systems is based on an approximate notion of simulation relation, a way of obtaining feedback control laws.