基于模糊逻辑的过程运行状态识别方法研究与应用

Study on fuzzy logic based operation state identification method and its application

工业生产过程存在许多内在的风险,设备及传感器的故障或操作条件不当都会导致异常状况的发生。面对工业生产过程采集的大量生产运行数据,尤其是同时出现几个异常状况时,操作人员依靠经验难以及时有效地处理。自动化和计算机技术的迅速发展,为过程工业及时有效地处理大量数据提供了可能,通过处理这些数据可得到运行过程的完整信息。本文提出的基于模糊逻辑过程运行状态的最少证据辨识法可全面考虑与过程相关变量、参数对过程运行状态的影响,通过最少证据辨识法可准确判断过程所处的运行状态。在美国杜邦公司己二酸生产过程中的成功应用表明,该方法可及时有效地判断过程所处的运行状态,及时发现故障,其辨识准确率高达95%,确保了过程运行的安全性。

Operating processing plants is inherently risky due to potentially abnormal situations caused by equipment or sensor failures or out of control process conditions. The fast increasing amount of process data makes process operation difficult to handle in lime and effectively, especially when more than one alarm are triggered in the operation. The rapid development of automation and computer technologies makes it possible for process industry to retrieve a huge amount of process data, from which complete information on process operation can be obtained. The method of minimal evidence method proposed in this paper makes it possible to consider all possible factors related to process operation,and to identify the process operation state by fuzzy logically recognizing the different pattern associated.The case study on adipic acid process in Victoria,Texas,USA,shows that this method can effectively identify the process operation states,find abnormal situations with the correct response rate of 95 %, and insure the operation safety.