高炉异常炉况判断专家系统的设计与实现

探讨了高炉异常炉况判断专家系统的结构,讨论了从传感器数据中抽取高炉况特征的方法,时态黑板结构与时间知识表示,对于炉况特征的模糊量化和专家知识中的参数设定,系统采用机器学习方法自动调整,对于高炉异常炉况的操作方法,系统采用规划－动作－反应方法向用出动国作建议,并对所动作动作建议实现实时的重构解释。

The Design and Realization of the Decision System for the Blast Furnace Abnormal State

The structure of blast furnace abnormal furnace state is approachad learning. The method of extracting the feature of blast furnace states from the sensor data,the temporal blackboard and the time knowledge expression,are discussed. The system applies the method of machine learning to adjust autmatically the fuzzy quantifization of blast furhace feature and the parameter assignment of the expert knowledge. As for the operating method of blast furnace abnormal furnace states,the operating suggestion is put forward through a "plan operate response" method and a realtime reconstructive explanation is given to the operating suggestion.