多传感器信息融合的燃气轮机状态监测系统

Multi-sensor Information Fusion of Gas Turbine Condition Monitoring System

为了克服单源传感器在燃气轮机状态监测与故障诊断中存在的局限性,在研究基于多传感器信息融合原理的基础上,开发一种燃气轮机状态监测及故障诊断系统。利用系统设置的多传感器信息融合结构模型,对多源传感器采集的信息进行预处理及特征征兆提取,并应用故障诊断综合评判规则对所有提取的有效信息进行故障决策。通过对某海上石油作业区的燃气轮机发电机组进行运行状态测试,获取比单源传感器系统更为可靠和全面诊断结论,具有更强的工程应用价值。

In order to overcome the limitation of monophyletic sensor in the gas turbine condition monitoring and fault diagnosis,to develop a kind of gas turbine condition monitoring and fault diagnosis system on base of multi-sensor information fusion principle.Using the system set of multi-sensor information fusion structure model,the multi-source sensors acquisition signs information preprocessing and feature extraction,fault diagnosis and applied to comprehensive evaluation rules for all extract effective information to make decision fault.Through the running state tests of gas turbine generator set in the offshore oil,the multi-sensor system is more reliable and comprehensive diagnosis than monophyletic sensor system,it have more stronger engineering application value.