论文部分内容阅读
目前磨矿系统的故障诊断主要依靠工人的经验,而这为故障诊断增加了大量不确定性。此外,磨矿系统的数据较为复杂,不仅工人难以对故障的发生进行准确判断,传统机器学习算法也由于数据的线性不可分而表现不佳。为了解决线性不可分的问题,我们使用神经网络进行故障分类;面对故障数据的高复杂度,为提高神经网络的表达能力,我们使用自动编码器增加网络深度;为减弱深层网络带来的过拟合现象,引入Drop Out降噪自编码,最终Drop Out降噪自编码网络对于故障的分类准确率达到90.4%。
At present, the fault diagnosis of the grinding system mainly depends on the experience of the workers, which adds a great deal of uncertainty to the fault diagnosis. In addition, the grinding system data is more complex, not only workers difficult to accurately determine the occurrence of the failure, the traditional machine learning algorithm because of linear data can not be divided and poor performance. In order to solve the problem of linear inseparability, we use neural network to classify the faults. In the face of the high complexity of fault data, we use the automatic encoder to increase the network depth in order to improve the performance of neural network. Combined with the phenomenon, the introduction of Drop Out noise reduction self-encoding, the final Drop Out noise reduction self-coding network for fault classification accuracy rate of 90.4%.