摘要
引言
由于具有适用范围广、结构简单等优点,离心泵在化工生产中得到广泛应用。汽蚀是离心泵运行中的常见问题。汽蚀的产生和发展不仅影响流道内速度分布,使泵工况变坏、效率降低,而且影响其动态响应,长时间的汽蚀还可能严重损伤叶轮等过流部件[1]。寻找适当的故障诊断、识别方法,
Centrifugal pumps are widely used in chemical processes.Cavitation in pumps is one of the major causes leading to reduced efficiency.A diagnostic method of cavitation inception was put forward based on low frequency and high frequency characteristics of pump inlet pressure fluctuation signals.The marginal spectrum was obtained through empirical mode decomposition（EMD）of experimental data and Hilbert-Huang transform（HHT）.By qualitative analysis root-mean-square,and marginal spectrum band energy of each intrinsic mode function could be used for cavitation recognition.It took too much time to recognize when the characteristic dimension was high,therefore it was necessary to quantitatively analyze for simplication.A four-dimensional feature vector was put into the back propagation neural network for training and simulation,with the first and second level root-mean-square energy values of intrinsic mode function obtained through EMD as high frequency feature and the 0 to 20 Hz and 20 Hz to 40 Hz band energy values of marginal spectrum as low frequency feature.The method mentioned above increased the recognition rate by 7.26% and 3.59% with simulation time decreased by 77.72% when contrasting with wavelet analysis method and EMD energy entropy method.It had a strong influence on the training of network that 3 to 9 level energy entropy of experimental data EMD varied little with cavitation conditions.So EMD energy entropy method took much time of simulation with low recognition rate.By removing the redundant characteristics recognition rate increased by 3.59%,simulation time decreased by 77.72%.The feature of wavelet analysis method was more varied in different flow rates,therefore recognition rate was lower for different flow rates.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2012年第2期545-550,共6页
CIESC Journal
基金
吉林省教育厅'十一五'科学技术研究项目(200747)~~
作者简介
周云龙（1960-），男，教授。
联系人：刘永奇。
