A Novel Deep Model with Meta-Learning for Rolling Bearing Few-Shot Fault Diagnosis

Machine learning,especially deep learning,has been highly successful in data-intensive applications;however,the performance of these models will drop significantly when the amount of the training data amount does not meet the requirement.This leads to the so-called few-shot learning(FSL)problem,which requires the model rapidly generalize to new tasks that containing only a few labeled samples.In this paper,we proposed a new deep model,called deep convolutional meta-learning networks,to address the low performance of generalization under limited data for bearing fault diagnosis.The essential of our approach is to learn a base model from the multiple learning tasks using a support dataset and finetune the learnt parameters using few-shot tasks before it can adapt to the new learning task based on limited training data.The proposed method was compared to several FSL methods,including methods with and without pre-training the embedding mapping,and methods with finetuning the classifier or the whole model by utilizing the few-shot data from the target domain.The comparisons are carried out on 1-shot and 10-shot tasks using the Case Western Reserve University bearing dataset and a cylindrical roller bearing dataset.The experimental result illustrates that our method has good performance on the bearing fault diagnosis across various few-shot conditions.In addition,we found that the pretraining process does not always improve the prediction accuracy.

Human resources allocation for aircraft maintenance with predefined sequence

There are always large-scale items in the maintenances schedule of aircraft system, many of which have been fixed to be done in predefined sequences, which leads the workflow to be sys-tematically complex and makes this kind of problem quite different from all sorts of existing job-selection modes. On the other hand, the human resources are always limited and men have different working capabilities on different items, which make the allocation operation of human resources be much roomy. However, the final total time span of maintenance is often required to be as short as possible in many practices, in order to suffer only the lowest cost of loss while the system is stopping. A new model for op-timizing the allocation if aircraft maintenance human resources with the constraint of predefined sequence is presented. The ge-netic algorithm is employed to find the optimal solution that holds the shortest total time span of maintenance. To generate the ul-timate maintenance work items and the human resource array, the sequences among all maintenance work items are considered firstly, the work item array is then generated through traversal with the constraint of maintenance sequence matrix, and the human resources are finally allocated according to the work item array with the constraint of the maintenance capability. An example is demonstrated to show that the model and algorithm behave a satisfying performance on finding the optimal solution as expected.

Design of a novel mitochondria targetable turn-on fluorescence probe for hydrogen peroxide and its two-photon bioimaging applications

Considering that hydrogen peroxide(H2O2)plays significant roles in oxidative stress,the cellular signal transduction and essential biological process regulation,the detection and imaging of H2O2 in living systems undertakes critical responsibility.Herein,we have developed a novel two-photon fluorescence turn on probe,named as Pyp-B for mitochondria H2O2 detection in living systems.Selectivity studies show that probe Pyp-B exhibit highly sensitive response toward H2O2 than other reactive oxygen species(ROS)and reactive nitrogen species(RNS)as well as biologically relevant species.The fluorescence colocalization studies demonstrate that the probe can localize in the mitochondria solely.Furthermore,as a bio-compatibility molecule,the highly selective and sensitive of fluorescence probe Pyp-B have been confirmed by its cell imaging application of H2O2 in living A549 cells and zebrafishes under the physiological conditions.