A Bayesian compressive sensing

Stochastic harmonic function based random field simulation method incorporating with Gaussian mixture model to consider fusion of multi-groups of spatial data

Authored by

Jingran He, Ruofan Gao, De Cheng Feng, Michael Beer

Abstract

AbstractIn engineering applications, random fields provide a robust probabilistic framework for characterizing properties that exhibit spatial and temporal variations. In practice, however, data sets for random field modeling are often collected sequentially. Consequently, rationally fusing these multiple data sets requires further investigation to maximize data utilization. Previous studies have demonstrated the excellent performance of the Bayesian compressive sensing – stochastic harmonic function method in representing random fields based on a single set of data. In this study, this method is combined with a Gaussian mixture model to further extend its capability to represent random fields using multiple data sets. Specifically, the high-dimensional probability distribution of the wavenumber spectral density function can be obtained using the Gaussian mixture model. As a result, the convergence property of the proposed method is verified, and the method is applied in some engineering problems to validate its performance.

Details

Organisation(s): Institute for Risk and Reliability
External Organisation(s): Guangdong University of Technology
Tongji University
Jinan University
Southeast University (SEU)
University of Liverpool
Type: Article
Journal: Mechanical Systems and Signal Processing
Volume: 251
ISSN: 0888-3270
Publication date: 01.05.2026
Publication status: Published
Peer reviewed: Yes
ASJC Scopus subject areas: Control and Systems Engineering, Signal Processing, Civil and Structural Engineering, Aerospace Engineering, Mechanical Engineering, Computer Science Applications
Electronic version(s): https://doi.org/10.1016/j.ymssp.2026.114199 (Access: Closed )