论文部分内容阅读
为了建立蛋壳破损检测模型,试验采用敲蛋装置与声检测控制器组成的计算机蛋破损检测系统获取被检蛋声音信号数据。通过对蛋声音信号的功率谱分析而得到反映蛋壳破损特征的参数:功率谱面积的平均值x1,最大功率谱面积与最小功率谱面积的差值x2,X轴方向上质心的平均值x3,X轴方向上质心最大值与其最小值的差x4,Y轴方向上质心的平均值x5,Y轴方向上质心最大值与其最小值的差x6,共振峰频率的最大值x7和共振峰频率的最大值与其最小值间的差值x8;再通过Bayes原理建立与蛋壳破损特征参数相关的蛋壳破损模型。检验结果表明模型准确率达到92%。
In order to establish a broken shell detection model, the experiment uses a computer egg damage detection system composed of a knocking egg device and an acoustic detection controller to obtain the data of the tested egg sound signal. Through the analysis of the power spectrum of the egg sound signal, the parameters reflecting the characteristics of the broken eggshell are obtained: the average of the power spectrum area x1, the difference x2 between the maximum power spectrum area and the minimum power spectrum area, the average x3 , The difference x4 between the maximum value of the centroid and the minimum value thereof in the X-axis direction, the average value x5 of the centroid in the Y-axis direction, the difference x6 between the maximum centroid value in the Y-axis direction and the minimum value thereof, the maximum value x7 of the formant frequency and the formant frequency And the difference between its maximum value and its minimum value x8. Then, the Bayesian principle is used to establish the damage model of the eggshell which is related to the characteristic parameters of the damaged eggshell. Test results show that the model accuracy rate of 92%.