首页> 外文会议>CIRP Conference on Industrial Product-Service Systems >Gas Turbine Exhaust System Health Management Based on Recurrent Neural Networks
【24h】

Gas Turbine Exhaust System Health Management Based on Recurrent Neural Networks

机译:燃气轮机排气系统健康管理基于经常性神经网络

获取原文

摘要

Predictive maintenance of gas turbine exhaust system plays an important role in reducing the risk of failure and improving the safety of gas turbine operation. This paper proposed an effective method based on the recurrent neural networks (RNNs) and statistical process control chart to predict the condition of the gas turbine exhaust system. Firstly, the fuel flow and air flow of the input gas chamber are used to predict exhaust gas temperature by an RNN model. Secondly, an Xbar-s statistical process control chart is constructed via comparing the current exhaust gas temperature predicted by RNN with that measured by sensors. Then the chart can be used to monitor whether there are failures in gas turbine exhaust system and temperature sensors. This is the first time that recurrent neural networks are used to address the health management of gas turbine exhaust system. And the sensor failure is also taken into consideration. Finally, the proposed approach is evaluated in the numerical experiment.
机译:燃气轮机排气系统的预测维护在降低失效风险方面发挥着重要作用,提高了燃气轮机操作的安全性。本文提出了一种基于经常性神经网络(RNN)和统计过程控制图的有效方法,以预测燃气轮机排气系统的状况。首先,输入气室的燃料流量和空气流量用于通过RNN模型预测废气温度。其次,通过比较由传感器测量的RNN预测的当前排气温度来构造Xbar-S统计过程控制图。然后,图表可用于监视燃气轮机排气系统和温度传感器中是否存在故障。这是首次使用经常性神经网络来解决燃气轮机排气系统的健康管理。还考虑了传感器故障。最后,在数值实验中评估了所提出的方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号