论文部分内容阅读
日本钢管公司应用人工智能专门系统创制了高炉炉况在线、适时控制系统,并已用于该公司的福山厂5号高炉上。这种控制系统由不正常炉况的识别系统和炉温监控系统组成。通常,高炉炉况的在线控制是靠过程计算机来完成的,这种传统的计算机控制系统因受数模等的限制,对不正常炉况不可能作出快速响应。为了克服上述因难,日本钢管公司自1984年以来,一直在研究采用人工智能技术预测不正常高炉炉况的系统。这套新系统有两方面的功能:①初处理功能,即对温度计、压力计等探测器给出的数据进行初处理,排出变量模式;②参照存储的信息给出可靠的结论,提示操作者应采取的行动。这套系统作为高炉操作的保障措施,已令人满意地发挥着作用,在预告塌料方面已获得80％以上的良好效果。
Nippon Steel Pipe Co., Ltd. developed an on-line, timely control system for blast furnace conditions using artificial intelligence systems and has been used on the company’s Fukuyama Plant No. 5 blast furnace. This control system consists of an abnormal furnace identification system and a furnace temperature monitoring system. In general, the on-line control of the blast furnace condition is accomplished by the process computer, which is unable to respond quickly to abnormal conditions due to the limitations of the digital control system. In order to overcome the above difficulties, Nippon Steel Corporation has been studying, since 1984, a system that employs artificial intelligence techniques to predict abnormal blast furnace conditions. The new system has two functions: ① initial processing function, that is, the thermometer, manometer and other detectors given the initial data processing, discharge variable mode; ② reference to the stored information to give a reliable conclusion, suggesting that the operator Action to be taken. This system has been satisfactorily used as a safeguard for blast furnace operations and has achieved over 80% good results in predicting slump.