摘要
提出了一种基于模型的化工管路泄漏故障诊断方法。利用动态模拟来监测管路流动过程,并在流量发生异常时及时进行故障诊断。诊断过程通过动态模型的在线参数估计完成,在实现模型校正的同时可以预测流动过程的变化趋势并判断是否存在故障。流动模型基于质量衡算和机械能衡算构建,并采用递归结构实现复杂管网的模拟。文中分析了简单管路和复杂管路应用实例,并讨论了对诊断结果起重要作用的影响因素。
Diagnosis of leakage faults in pipeline transportation is an important area in chemical processing, and is mostly carried out using process history based methods and knowledge based methods. In this paper we propose a novel leakage fault detection and diagnosis method, using dynamic simulation to monitor the fluid flow process and identify leakages when large flow rate deviations occur. The inner leakage parameters are continuously updated via on-line correction, allowing the flow trends to be monitored and the existence of malfunctions to be detected simultaneously. The flow model is based on the principles of mass balance and mechanical energy balance, and is simulated using a recursive solution. Case studies of the proposed method are presented for a simple pipeline and a series pipeline, and the effects of different factors on the results are analyzed.
出处
《北京化工大学学报（自然科学版）》
CAS
CSCD
北大核心
2008年第5期18-23,共6页
Journal of Beijing University of Chemical Technology(Natural Science Edition)
基金
教育部留学回国人员科研启动基金(2005-29)
山东省优秀中青年科学家科研奖励基金(2006BS05005)
关键词
故障诊断
管路泄漏
动态模拟
参数估计
fault diagnosis
pipeline leakage
dynamic simulation
parameter estimation
