Title :
Architectural models for intelligent machines
Author :
Valavanis, K.P. ; Saridis, G.N.
Author_Institution :
Northeastern University, Boston, MA
Abstract :
Intelligent Machines are defined to be hierarchical structures composed of three interactive levels: organization, coordination and execution. An architectural model that can accommodate fast and reliable operation for the levels individual functions is proposed and explained. The model is composed of hardware units including special memories, long-term, short-term and buffers, to store permanent and temporary information, memories to store probabilities, costs and terminal sets, and specific hardware to upgrade the stored information. The architecture proposed is suitable for the decision phase and upgrade phase of Intelligent Machines. The decision phase involves formulating complete and compatible plans to execute a user requested job. The upgrade phase involves the upgrade of the probabilities and costs associated with the best possible plan formulated during the decision phase.
Keywords :
Buffer storage; Concatenated codes; Costs; Hardware; Intelligent control; Intelligent structures; Intelligent systems; Machine intelligence; Random variables; Systems engineering and theory;
Conference_Titel :
Decision and Control, 1986 25th IEEE Conference on
Conference_Location :
Athens, Greece
DOI :
10.1109/CDC.1986.267578