Adopting wavelet neural network,fault forecast of characteristic parameters of mine ventilator was realized by forecasting time series of key characteristic parameters of ventilator that include vibration intensity.

Based on these measurements,a conclusion was drawn that small stiffness of the foundation was important reason of vibration intensity that was over criterion.

In the running course of large rotany sets,the vibration intensity of sets usually behaves asnonstationary time series with tendency characteristics.

Based on theoretic analysis,the arithmetic of vibration severity using DFT is investigated,and precautions in actual application are given in combination with practical experience.

After the measurement of vibration severity and vibration level difference of some kind of gearbox, the measuring results are presented in detail.

By imposing an excitement source to stimulate transformer winding vibrations, and by means of both the total vibration velocity level function and vibration velocity intensity index function, features of the winding\'s vibration signal changes can be found out, and based upon the findings the status of the winding in operation is determined.