跳到主要导航 跳到搜索 跳到主要内容

Accelerated algorithm for BTT identification parameter with GMC sparse regularization

  • Xi'an Jiaotong University
  • Thermal Power Research Institute

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

1 引用 (Scopus)

摘要

Accurate identification of vibration parameters from blade tip timing (BTT) undersampled signals is essential for rotating blade vibration monitoring. However, the traditional parameter identification method of BTT signal depends on the prior. The existing sparse regularization method underestimates the reconstructed signal amplitude and has low computational efficiency. This paper resorts to an accelerated algorithm for BTT identification parameters based on generalized minimax-concave (GMC) sparse regularization to accurately and quickly identify amplitude and frequency parameters from undersampled signals. For amplitude underestimation, the non-convex GMC penalty is introduced so that the sparsity of the estimation is improved, and the convexity of the cost function is preserved. Moreover, Nesterov's accelerated iterative computation strategy is resorted to rapidly improving the convergence performance of obtaining the global optimum. The simulation results show that by reconstructing the BTT signal, the presented parameter identification algorithm based on accelerated generalized minimax-concave (AGMC) improves the computational rate with the inherited merits of accuracy.

源语言英语
主期刊名2022 International Conference on Sensing, Measurement and Data Analytics in the Era of Artificial Intelligence, ICSMD 2022 - Proceedings
出版商Institute of Electrical and Electronics Engineers Inc.
ISBN(电子版)9781665492812
DOI
出版状态已出版 - 2022
活动3rd International Conference on Sensing, Measurement and Data Analytics in the Era of Artificial Intelligence, ICSMD 2022 - Harbin, 中国
期限: 22 12月 2022 → 24 12月 2022

丛书

姓名2022 International Conference on Sensing, Measurement and Data Analytics in the Era of Artificial Intelligence, ICSMD 2022 - Proceedings

会议

会议3rd International Conference on Sensing, Measurement and Data Analytics in the Era of Artificial Intelligence, ICSMD 2022
国家/地区中国
市Harbin
时期22/12/22 → 24/12/22

学术指纹

探究 'Accelerated algorithm for BTT identification parameter with GMC sparse regularization' 的科研主题。它们共同构成独一无二的学术指纹。

引用此