ESCORT: the application of causal knowledge to real-time process control
Proc. of the fifth technical conference of the British Computer Society Specialist Group on Expert Systems on Expert systems 85
A continuous real-time expert system for computer operations
IBM Journal of Research and Development
Intelligent decision aids for process environments: an expert system approach
Proceedings of the NATO Advanced Study Institute on Intelligent Decision Support on Intelligent decision support in process environments
Selection criteria for expert system shells: a socio-technical framework
Communications of the ACM
A Knowledge-Based Control Architecture with Interactive Reasoning Functions
IEEE Transactions on Knowledge and Data Engineering
Hi-index | 0.02 |
Extase, a situation assessment system that finds the causes of alarm signals relating to the heating furnace of a distillation column, is presented. The system is centered around two design principles. First, heterogeneous knowledge is represented by causally describing process functions. Second, a hypothesis network that supports local search control, stores time-dependent dynamic information, and forms a basis for the explanation facility is progressively built. The motivations and requirements for this research are examined, the process environment that served as an example is described, and system implementation is discussed. Extase has been installed and tested in a refinery and has completed its first development cycle. The authors describe and discuss system implementation principles in detail. Finally, they present the results of field testing and assess the system's limits and strengths.