No abstract

No abstract

No abstract

In [20], a proposal was made for constructing highly adaptable software based on an architecture that uses distributed agents. In this papel; a case study is described which follows this approach. The important contribution lies in developing a model of the application domain, in this case the relational database domain, which is then available fora domain agent to manipulate in developing strategies for changing the application software. The application domain is modelled as a set of entities related by constraints of different types. Domain adaptability is accomplished through both the creation of new entities and new constraints between entities. Why and how this approach is different from existing approaches to adapting databases is explained. Initial results have produced a domain model which makes explicit the potential points of adaptability in the domain being modelled, making it easier to adapt software. Additionally, the use of agent interaction protocols provides a basis for research into inter-operability between separate domains.

Several deterministic identification problems, with partial realization being a special case, are unified in the framework of the mathematical problem "generalized dynamic covers".An algorithm to find such a minimal dynamic cover as well as a uniqueness criterion is given, which yields several identifiability results. This problem also includes the "observer" and the ltexact model matching" problems, as well as the problem of finding "minimal inverses for linear systems with arbitrary initial states".