A Study on Joint Noise Reduction of Bridge Monitoring Data Using GJO-MVMD and Wavelet Thresholding
Xianglong Li, Lei Wang*
School of Civil Engineering and Architecture, University of Jinan, Jinan, China
* Corresponding Author: Lei Wang. Email: 
Structural Durability & Health Monitoring https://doi.org/10.32604/sdhm.2026.085601
Received 14 May 2026; Accepted 15 July 2026; Published online 12 August 2026
Abstract
To address the engineering challenges posed by the susceptibility of raw data collected by bridge health monitoring systems to interference from complex environmental noise in real-world applications, as well as the severe mode aliasing issues associated with Empirical Mode Decomposition when processing non-stationary signals, this paper proposes a joint denoising and feature separation technique that integrates the Golden Jackal optimization algorithm with multivariate variational modal decomposition and wavelet thresholding. Addressing the shortcomings of traditional variational modal decomposition—which relies heavily on manual parameter tuning—this framework introduces the Golden Jackal optimization algorithm. By strictly using envelope entropy as the fitness function, it performs adaptive, synchronized optimization of the decomposition order and penalty factors for multivariate variational modal decomposition, thereby eliminating the engineering risks and time costs associated with blind trial-and-error. Based on this, a secondary, refined noise-reduction process was applied to the mixed high-frequency modes using wavelet thresholding techniques. The method was validated by generating simulation signals and analyzing deflection-monitoring data from an in-service long-span cable-stayed bridge. The results indicate that this combined algorithm not only effectively filters out high-frequency fluctuations caused by random environmental disturbances but also accurately extracts the true low-frequency deformation trends driven by environmental factors such as temperature. Compared to traditional methods, this approach demonstrates significant advantages in quantitative metrics such as the noise reduction ratio and residual variance ratio. This study achieves high-fidelity separation of the structure’s true response signal from environmental noise, effectively eliminating baseline drift interference caused by temperature fluctuations, and provides a solid data foundation for the online assessment and dynamic early warning of bridge structural health.
Keywords
Bridge health monitoring; multivariate variational mode decomposition; golden jackal optimization
Open Access
Login
Register
Submit a Paper
Propose a Special lssue
Download PDF
Downloads
Citation Tools
