http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Frequency-Domain Data Augmentation of Vibration Data for Fault Diagnosis using Deep Neural Networks
Minseon Gwak,Seunghyun Ryu,Yongbeom Park,Hyeon-Woo Na,PooGyeon Park 제어로봇시스템학회 2022 제어로봇시스템학회 국제학술대회 논문집 Vol.2022 No.11
This paper proposes a data augmentation method for vibration data-based fault diagnosis using deep neural networks. The proposed method is devised to deal with the practical problem in applying trained models to facilities, where frequency-domain features of data vary according to the change in the working environment of the facilities. In the proposed method, training data are augmented by scaling the frequency-domain features of raw training data by small amounts generated by a normal distribution. The proposed method is implemented to preserve the symmetricity of the positive and negative frequency-domain components and return the real part of the complex inverse transformed data as final augmented data. The advantage of the proposed method is verified by simulation, where the operating conditions of training and test data differ. Moreover, it is shown that the proposed method can improve the accuracy of models better compared to a time-domain data augmentation using similar random scaling.
A filtered-x scheduled step-size active noise cancellation algorithm considering implementation
Taesu Park,Minseon Gwak,PooGyeon Park 제어로봇시스템학회 2021 제어로봇시스템학회 국제학술대회 논문집 Vol.2021 No.10
This paper proposes an active noise cancellation algorithm by using adaptive filter. We apply the normalized subband adaptive filter which is one of the adaptive filter algorithm to a filtered-x algorithm. Because the adaptive filter algorithm is applied in acoustic noise cancellation system, we rearrange the filter update recursion formula. We analyze the mean-square deviation of the NSAF to schedule the step size according to the iteration. As a result, a step size table was created, and the proposed algorithm shows similar performance to other variable step size NSAFs by changing the step size without additional online calculation. The generated step-size table can be modified online in proportion to not only the number of taps but also the number of subbands. The noise cancellation simulation shows that the proposed algorithm performs better than the existing active noise cancellation algorithm using variable step size NSAF without additional calculation. The simulation results also show that the proposed algorithm is robust even under erroneously estimated environmental conditions.