Degradation Feature Extraction Method for Prognostics of an Extruder Screw Using Multi-Source Monitoring Data.

Laboratory-scale data on a component level are frequently used for prognostics because acquiring them is time and cost efficient. However, they do not reflect actual field conditions. As prognostics is for an in-service system, the developed prognostic methods must be validated using real operational data obtained from an actual system. Because obtaining real operational data is much more expensive than obtaining test-level data, studies employing field data are scarce. In this study, a prognostic method for screws was presented by employing multi-source real operational data obtained from a micro-extrusion system. The analysis of real operational data is more challenging than that of test-level data because the mutual effect of each component in the system is chaotically reflected in the former. This paper presents a degradation feature extraction method for interpreting complex signals for a real extrusion system based on the physical and mechanical properties of the system as well as operational data. The data were analyzed based on general physical properties and the inferred interpretation was verified using the data. The extracted feature exhibits valid degradation behavior and is used to predict the remaining useful life of the screw in a real extrusion system.

Fault Diagnosis Method for Human Coexistence Robots Based on Convolutional Neural Networks Using Time-Series Data Generation and Image Encoding.

This paper proposes fault diagnosis methods aimed at proactively preventing potential safety issues in robot systems, particularly human coexistence robots (HCRs) used in industrial environments. The data were collected from durability tests of the driving module for HCRs, gathering time-series vibration data until the module failed. In this study, to apply classification methods in the absence of post-failure data, the initial 50% of the collected data were designated as the normal section, and the data from the 10 h immediately preceding the failure were selected as the fault section. To generate additional data for the limited fault dataset, the Wasserstein generative adversarial networks with gradient penalty (WGAN-GP) model was utilized and residual connections were added to the generator to maintain the basic structure while preventing the loss of key features of the data. Considering that the performance of image encoding techniques varies depending on the dataset type, this study applied and compared five image encoding methods and four CNN models to facilitate the selection of the most suitable algorithm. The time-series data were converted into image data using image encoding techniques including recurrence plot, Gramian angular field, Markov transition field, spectrogram, and scalogram. These images were then applied to CNN models, including VGGNet, GoogleNet, ResNet, and DenseNet, to calculate the accuracy of fault diagnosis and compare the performance of each model. The experimental results demonstrated significant improvements in diagnostic accuracy when employing the WGAN-GP model to generate fault data, and among the image encoding techniques and convolutional neural network models, spectrogram and DenseNet exhibited superior performance, respectively.