Small Short-sound Drill Dataset
收藏资源简介:
本研究使用的是小型短声钻数据集,由瑞典Valmet AB创建,用于监测钻机故障。数据集包含三类声音:异常声、正常声和无关声,总计约41,250条声音记录。数据集通过四台AudioBox iTwo Studio麦克风在96kHz采样率下录制。创建过程中,通过数据增强技术增加了数据集的样本量,以提高深度学习模型的性能。该数据集主要应用于工业机器故障检测,旨在通过声音分析提前识别钻机故障,减少机器停机时间和维护成本。
The small-scale short sound drilling dataset used in this study was created by Valmet AB of Sweden for drilling machine fault monitoring. This dataset contains three categories of sounds: abnormal sounds, normal sounds, and irrelevant sounds, with a total of approximately 41,250 audio recordings. It was recorded using four AudioBox iTwo Studio microphones at a sampling rate of 96 kHz. During its creation, data augmentation techniques were applied to expand the dataset's sample size so as to improve the performance of deep learning models. This dataset is primarily applied to industrial machine fault detection, aiming to identify drilling machine faults in advance through sound analysis, thereby reducing machine downtime and maintenance costs.




