论文部分内容阅读
铁塔加工过程中需要对角钢进行夹紧、冲孔、剪切等操作,不同材质的角钢加工时会发出不同的声音,因此提出利用声音信号检测Q235和Q345两种常用角钢材质。利用传感器和采集仪搭建试验系统,采集声音信号的连续波形,并提取单个冲孔周期声音信号的波形作为后续判别的依据。针对单个周期的声音信号,提取美尔频率倒谱参数(MFCC)作为特征参数,并利用动态时间规整(DTW)方法计算待测模板与Q235和Q345两种标准模板之间的距离,距离小者判定为该种角钢材质。试验分别采集了4种型号角钢的2组样本,验证了上述方法识别材质的有效性。
During the processing of the tower, the angle steel needs to be clamped, punched, sheared and the like. When different angles of steel are processed, different sounds will be emitted. Therefore, it is proposed to use the sound signal to detect Q235 and Q345 two commonly used angle steel materials. Using the sensor and the collector to build a test system, a continuous waveform of the sound signal is collected, and the waveform of the sound signal of a single punching cycle is extracted as the basis for subsequent discrimination. For a single cycle of the sound signal, the frequency of MFCC is extracted as a feature parameter, and the dynamic time warping (DTW) method to calculate the distance between the template and the two standard templates Q235 and Q345, the distance Judged as the kind of angle steel material. Two sets of samples of four types of angle steel were collected and validated the effectiveness of the above method to identify the material.