Track:
All Papers
Downloads:
Abstract:
Our research in intelligent control systems has led to the design of a unique software architecture that combines heuristic and knowledge-based methods with traditional approaches to control. Emulation of the complex procedures used by human experts in controlling particle accelerators requires a hybrid architecture integrating methodologies for planning, intelligent search, and pattern recognition, among others. This system is distributed and hierarchical to utilize parallel problem-solving in the face of time sensitive control applications and to decompose complex control problems into more manageable subtasks. We use an object-oriented abstraction layer and a component-based architecture for high portability.