Unsteady noise test technology and time-frequency refinement analysis methods were used to investigate the cutting noise characteristics of the larger cold saw. The noises in different time-intervals in sawing process were tested. Then, by eliminating the background noise, the cutting noise characteristics were analyzed by the time-frequency refinement analysis method. The results show that the sawing sound field has higher sound pressure level, the radiation frequency has stronger time-varying performance, the sound energy is focused on the lateral direction of the cutting, and high-frequency is very obvious in the noise. This work may provide a technical support for noise test, sound field characteristics analysis and noise control of some complex machines such as fans, compressors, etc.

The auto-regressive(AR) model of exhaust noise in diesel engine’s seven operating modes about the valve train was established. The auto-regressive parameters were regarded as the characteristic vectors to build the fault diagnosis model which was based on extreme learning machine. The proposed method was also compared with that of BP neural network, RBF neural network and support vector machine. Experimental results show that exhaust noise can be used for fault diagnosis in diesel engine’s value train. The classification accuracies of above four different intelligent methods are all higher than 95%. But the method of extreme learning machine has advantage apparently in training rate.