Continuous Casting Slag Detection Expert System Based on CLIPS

Abstract

In order to resolve the pouring state recognition problem of continuous casting slag detection system (CC-SDS) and reduce its control delay, a kind of C language integrated production system (CLIPS) based CC slag detection expert system (SDES) was brought forward. Multi-channel real-time vibration signal collection system oriented to slag detection was established, and it could supply stable real-time support for the expert system. According to factual investigation results of CC production field, system knowledge base and fact base were realized by CLIPS frame. Using application program interface (API) of CLIPS, inference engine of the system was set up with relative transcendent production rule and detection algorithms in Linux environment. Combined Protégé technology, system developing period was decreased, and the interactivity performance was strengthened during the course of system development. Based on modularization principle, the reduced and configurable light graphic components library was designed and completed by Qt/E, a mature graphic user interface (GUI) tool library, and it could supply Chinese operating environment. Industry field experiments prove that this system has good cross-platform ability, can effectively judge the typical slag state, and can supply corresponding operating advices and its confidence degree.