A Hybrid KF-RDT-SSI Framework for Operational Modal Analysis of Offshore Jacket Platforms: Full-Scale Field Evaluation

Operational modal analysis (OMA) is important for continuous structural health monitoring (SHM) of in-service offshore jacket platforms. In practice, long monitoring records increase the computational cost of SSI, while environmental and measurement interference can affect the recorded response. This study applies a hybrid procedure combining Kalman filtering (KF), the Random Decrement Technique (RDT), and Stochastic Subspace Identification (SSI). KF is used as a fixed-parameter preprocessing step, RDT extracts shorter free-decay signatures, and SSI is used to identify the natural frequency. The procedure is examined using a finite element (FE) model, a laboratory-scale model, and full-scale measurements from the same operating offshore platform represented by the FE model. For a representative 50 Hz record, RDT reduced the data entering SSI from 38,553 to 1200 samples (96.89%) and reduced the number of Hankel-matrix elements by 97.05%. Under identical SSI settings, the complete KF-RDT-SSI procedure required about one-fifth of the Standard SSI runtime in the benchmark environment. In the four-channel laboratory test, the dominant spectral peak remained at 6.1279 Hz before and after KF processing. KF-RDT-SSI identified 6.1204–6.1353 Hz, with a mean absolute relative deviation of 0.095% from the FFT result. The field measurements identified the fundamental natural frequency at approximately 1.56–1.57 Hz under transient and continuous operating conditions. The full-scale measurements provide the main evidence for the field applicability of the procedure. A systematic assessment under controlled noise types and signal-to-noise ratios is left for future work.