摘要
针对推拉式酸洗机组自动穿带过程中因来料厚度、平直度等差异而造成的堆钢问题,研究了一种堆钢预报警系统,其具有自动穿带、带头跟踪和堆钢预报警3项功能。介绍了该预报警系统的逻辑判断模型。为提高预报警准确率,引入BP神经元网络模型对预报警系统进行了优化。该预报警系统投入使用3年,准确率达到97%。
For the problem of piece piled during the automatic threading process caused by the difference of thickness and straightness in the push-pull pickling unit,a pre-alarm system was developed,which has three functions of automatic threading,strip head tracking and piece piled pre-alarm.The logical judgment model of the pre-alarm system was introduced.In order to improve the accuracy of pre-alarm,the BP neural network model was applied to optimize the pre-alarm system.The pre-alarm system had been put into use for 3 years with an accuracy rate of 97%.
作者
耿运祥
GENG Yun-xiang(Beijing Research Institute of Automation for Machinery Industry Co.,Ltd.,Beijing 100120,China)
出处
《轧钢》
2019年第5期70-73,共4页
Steel Rolling
关键词
自动穿带
带头跟踪
堆钢
预报警
神经元网络模型
automatic threading
strip head tracking
piece piled
forecast and alarm
neural network model