Hybrid compression scheme based on VMD optimization algorithm application to mechanical equipment monitoring

• Published in: The Journal of supercomputing Vol. 80; no. 4; pp. 5341 - 5362
• Main Authors: Song, Liqiang, Wang, Huaiguang, Yang, Baojian
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.03.2024; Springer Nature B.V
• Subjects: Algorithms; Bandwidths; Compilers; Computer Science; Data compression; Data transmission; Decomposition; Entropy; Fault diagnosis; Fourier transforms; Interpreters; Monitoring; Optimization; Optimization algorithms; Parameters; Processor Architectures; Programming Languages; Signal to noise ratio; Software; Spectrum allocation; Wavelet transforms; Lossless compression; Variational mode decomposition; Intrinsic mode function components; Equipment monitoring; Lifting wavelet transform
• ISSN: 0920-8542; 1573-0484
• DOI: 10.1007/s11227-023-05663-9

The traditional variational mode decomposition (VMD) algorithm causes under-decomposition and over-decomposition problems when performing signal decomposition. Many studies have addressed various optimization algorithms. When the optimization parameters are set too small, the best decomposition cannot be found. On the other hand, setting too large parameters leads to a waste of computational resources. Therefore, an improved VMD algorithm is proposed in this paper, which effectively solves the above problems by establishing an adaptive method for choosing the values of decomposition parameters and its iteration termination condition. In order to meet the monitoring data transmission requirement, a hybrid compression scheme based on the improved VMD is proposed. The monitoring signal is effectively decomposed by the improved VMD algorithm, and then thresholded by the lifting wavelet transform (LWT) method, and further compressed by combining quantization as well as entropy encoding methods. It is experimentally verified that its compression scheme can effectively improve compression performance and has a high signal-to-noise ratio, providing an effective solution for data transmission in equipment monitoring.