【摘要】 在大批量小型电机生产线上,普遍采用的听音质检工序需要大量训练有素的工人。针对电机质检工序声音信号的统计特征及质检工艺的特点,提出基于一类学习的异响检测的方法。该方法以正常样本为基础建立质检判别函数,避免了其他分类算法要求训练样本类别全面和覆盖广泛的条件。给出了该方法的实现过程,并通过大量实测样本验证了该方法的有效性。更多还原

【Abstract】 In production lines of a large number of small motor,listening quality inspection procedure widely used needs a lot of well trained workers.According to the statistical characteristics of the sound signal in motor quality inspection process,a method based on one class of learning is presented to detect abnormal sound.With the foundation of normal samples,a discriminant function of quality inspection is established,which avoides to many conditions that comprehensive and extensive training samples required in other classification algorithms.The implementation of the method is presented,and the proposed method is validated by a large number of measured samples.更多还原