Recurrent convolutional neural network: A new framework for remaining useful life prediction of machinery

Machine health monitoring has become increasingly important in modern manufacturers because of its ability to reduce downtime of the machine and cut down the production cost. Enormous signals acqui...

A novel ResNet-based model structure and its applications in machine health monitoring:

Application of machine vision for tool condition monitoring and tool performance optimization–a review

The scarcity of abnormal data leads to imbalanced data in the field of monitoring tool wear conditions. In this paper, a novel multi-index generative adversarial network (MI-GAN) is proposed to detect the tool wear conditions subject to imbalanced signal data. First, the generator in the MI-GAN is trained to produce fake normal signals, and the discriminator computes scores of testing signals and generated signals. Next, the generator detects abnormal signals based on the performance of imitating testing signals, and the discriminator will compute the scores of testing signals and generated signals. Subsequently, two indexes, i.e., - norm and temporal correlation coefficient (CORT), are put forward to measure the similarity between generated signals and testing signals. Finally, our decision-making function further combines - norm and CORT with two discriminator scores to determine the tool conditions. Experimental results show that our method obtains 97% accuracy in tool wear detection based on imbalanced data without manual feature extraction, which outperforms traditional machine learning methods.

/pdf/a-multi-index-generative-adversarial-network-for-tool-wear-19mnh4iea9.pdf

A Multi-Index Generative Adversarial Network for Tool Wear Detection with Imbalanced Data

Measurement of cemented carbide-PCD microdrill geometry error based on computer vision algorithm

In modern manufacturing systems and industries, an increasing number of researches which are absorbed in recognition of abnormal process, apply data-driven model to analyze the manufacturing process. However, considering the sequence data is unable to be imported in the regression model and classification model, researchers need to investment most of time in feature extraction methods to recognize the abnormal process. In the last few years, with the development of neural network algorithm deep learning methods, which redefine the processing way from raw data, the neural network has been used to address raw sensory data. In this study, which is based on the multi-source perception signal of machining process, the CNNs algorithm's convolutional layer extracting information features, the BN normalizing retention feature information and the pooling layers reducing the feature matrix contribute to the establishment of a state prediction model which could predict and identify the degree of wear of the tool and the abnormal state of the machining process. After that, a tool wear test is introduced to comparing results of the traditional signal processing method with CNN algorithm and the feasibility of the CNN algorithm is demonstrated.

Jianglin Xiao

Papers

Detection for Cutting Tool Wear Based on Convolution Neural Networks