摘要
根据离心压缩机转子所承受的、可实时监测的轴向力与级密封间隙的关系,提出了基于轴向力的离心压缩机级密封状态的预测方法.在分析叶轮两侧缝隙内泄漏气体流动的基础上,采用计算流体动力学方法对其进行仿真,得到轮盘及轮盖上的压力分布,由此计算出轴向力.把不同级密封间隙值及对应的轴向力作为输出、输入向量,对BP神经网络进行训练,以获得在线状态预测模型.测试结果表明,该模型对离心压缩机级密封的状态预测是有效的.
According to the analysis on the relationship between the axial thrust acting on the impeller and be- ing monitored in real time and the clearance value of the stage seal of centrifugal compressor, an approach to predict the condition of the stage seal of centrifugal compressor based on the axial thrust is presented. The gas flow in the gap outside of the impeller is analyzed and simulated with CFD technology, then the pressure distri- butions on the wheel disc and impeller shroud are obtained. In succession, the axial thrust can be calculated. 24 clearance values of the stage seal and the corresponding axial thrusts are taken as output and input respec- tively to train BP neural network, and a condition predicting model of the stage seal of centrifugal compressor is obtained. Test results show that the model is effective for predicting the condition of the stage seal of centrifugal compressor.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2009年第11期25-28,共4页
Journal of Harbin Institute of Technology
基金
国家自然科学基金资助项目(10772061)
黑龙江省自然科学基金重点项目(ZJG0704)
关键词
离心压缩机
级密封
状态预测
轴向力
BP神经网络
centrifugal compressor
stage seal
condition prediction
axial thrust
BP neural network
