Detecting steel material based on sound signal

For the processing of the pylon, there is need do some operations on it, such as clamping, punching, shearing and other percussion. It will produce noise when shower punching on the angle cleat. Then we put forward the proposal that using the voice signal to test steel material. Q235 and Q345 are the general material. Using sound sensor and collection device structures to put up experimental system. Sound signal has been collected and the standard template library has been established according to the system. To the actual processing of the collected test sound signal, Mel frequency cepstrum parameters (MFCC) has been extracted. The Mel frequency cepstrum parameters are that the characteristic parameters. Then calculates the distance between the measured voice signal and standard templates using Dynamic Time Warping DTW method. The test results show that the system can accurately detect different models of two kinds of Angle steel material.