摘要
针对石化生产过程的高危性,开发了石化过程在线故障监测系统。通过OPC(OLE for process control)接口从生产现场采集实时数据,采用BP神经网络(back-propagation artificial neural network,BPNN)的模式识别方法,对生产过程进行实时故障监测,及时发现故障工况并提示操作人员采取相应措施,以减小系统运行的风险。BP神经网络的训练数据来自历史数据库,用户根据已发生过的故障工况确定训练数据的时间范围。BP网络模型的各项参数根据多次试验得到。对某工段的10个故障,其故障诊断准确率达到90%以上,具有较高的实时性和准确性。
For high-risks are in petrochemical production process,it developed an on-line fault monitoring system for petrochemical process in this paper.Through the OPC（OLE for process control）interface,it collected real-time data from the production field,and monitored the process by using pattern recognition method of back-propagation neural network（BPNN）.The system monitored the production process in real-time,detected fault conditions timely and prompted the operator to take measures to reduce the risk of system operation.The training data for BPNN came from the historical database,which the user could determine the time range of it from the fault conditions had already occurred.Parameters of BPNN model obtained from repeated experiments.For the 10 fault of one petrochemical section,the accuracy rate of the fault diagnosis is 90%.The application indicated that the system had a high real-time performance and accuracy.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2010年第10期1418-1420,共3页
Computers and Applied Chemistry
基金
国家高技术研究发展计划(863)(2009AA04Z133)
