A.D. Bogobowicz (Canada)
process recognition, non-linear hyperbolic systems, measure theory, entropy production.
In the paper some problems of distributed parameter systems are considered from the point of view of object oriented methods, which are still lacking theoretical background. A canonical object is formulated for approximation of the phase space of non-linear distributed parameter systems. The purpose of the approach is to create faster algorithms that solve a system local model. Migrating canonical objects are specified to enable process recognition in the open domain of the system. In the paper we are primarily interested in controlled processes and the introduced invariant measure chosen for migrating objects allows us to identify local properties and choose a unique local value. As an example the process described by a non-linear hyperbolic system is considered. The local behaviour is approximated by a parabolic system.
Important Links:
Go Back
