Coupling an expert system with industrial real-time processes

A diagnostic real-time expert system, applied to an industrial chemical process, is used as a testbed for research on integration of symbolic and numeric computations in a unified framework. The algorithms reveal the precise role and desirable characteristics of synergism between statistical process control and the expert system. Inferencing on the process performance is founded on the consistent and continuous acquisition of the process parameters, statistical and mass balance calculations, heuristic rules processing, and history data reporting. The implementation status and runtime performance of the system are given.