Experimental research on turbine rotor vibration fault diagnosis based on fractal correlation dimension
According to the common fault of steam-turbine vibration, rubbing, loosing, uncountershaft and mass unbalance can be simulated on the Bendy experiment table. With the theory of phase space reconstruction, vibration signal, a kind of random fractal that meets the statistic self-similarity, was reconstructed With the calculation of the fractal dimension, the dynamical character of the system can be reproduced. The results show that die rubbing fault wave pattern is the most disordered, and the concerned correlation dimension number is the largest, especially the metal rubbing correlation dimension number is 1.97. The correlation dimension of loosing fault is a little bit lower and the mass unbalance fault correlation dimension is the smallest, namely, 1.58. Therefore, the correlation dimension can be used as a kind of vibration fault sign in steam turbine safety. The calculated results also shows that different types of faults must get different correlation dimension increments, the increments of rubbing and loosing situation are larger, and the correlation dimension of mass unbalance fault situation has seldom difference with its initial situation. Therefore, the relative difference between the fault and the initial situation can show the type of fault to a certain extent.
Ping Liang Xin-feng Long Fu-mei Fan
School of Electric Power, South China University of Technology, Guangzhou 510640, China School of Chemical Engineering and Energy, South China University of Technology,Guangzhou 510640, Ch School of Electric Power, South China University of Technology, Guangzhou 510640, China;Power Engine
国际会议
The 5th International Conference on Nonlinear Mechanics(第五届国际非线性力学会议)
上海
英文
1299-1304
2007-06-11（万方平台首次上网日期，不代表论文的发表时间）