マルチラベルディープニューラルネットワークを用いた機械部品製造装置の異常検知

抄録

In this paper, we discuss on fault detection method for a machine parts manufacturing equipment. For the fault detection, a machine learning technique is employed based on operating time data which is relatively easy to obtain. An issue to be solved is that it is not easy to obtain fault data to be detected by daily normal operation of the equipment. It is required to apply unsupervised learning techniques or prepare appropriate fault data for machine learning. In this paper, the fault data were generated from the distribution of the normal data and a multi-label deep neural network(ML-DNN) is applied to detect faults in multiple actions simultaneously. Some results show the effectiveness of the proposed procedure.