Publications by authors named "Gonghai Wang"
With the rapid development of industrialization and urbanization, the impact of noise on people's health has become an increasingly serious issue, but it is still a challenge for the reducing the noise due to its complex property. Textiles with many loose porous structures have gained much significant attentions, thus chenille yarns with plush fibers on the surface, and polyester monofilament were chosen to fabricate the integrated knitting yarns, and their fundamental and mechanical properties were fully evaluated. The results showed that the diameter and braiding angle of the blended yarns decreased with the increase of pitch, resulting in a linear correlation of R > 0.
View Article and Find Full Text PDFArticle Synopsis
- Deep learning is becoming popular for health management and maintenance in rotating machinery, but challenges like domain shifts and lack of labeled training data limit its effectiveness.
- The study introduces a domain adaptive framework called deep Multiscale Conditional Adversarial Networks (MCAN) that includes a shared feature generator with an attention mechanism and two domain classifiers based on BiLSTM for better machinery fault diagnosis.
- Evaluations using various datasets show that MCAN significantly outperforms existing methods in transferability and stability, marking an important advancement in the use of deep learning for machinery maintenance.