Load Inversion Method for Jacket Platform Structures Based on Strain Measurement Data

Due to the difficulty of directly measuring external loads on jacket platform structures and the challenges in accurately expressing them through analytical formulas, this study proposes a load inversion method based on local strain measurement data to obtain the time–history curves of structural loads. The method establishes a mapping relationship between unknown loads and measured strains based on the quasi-static superposition principle. An Improved Sine Cosine Algorithm, combined with an Opposition-Based Learning, is introduced to optimize the placement of strain sensors. The unknown loads are solved using a least squares approach integrated with Tikhonov regularization. The method was validated through indoor loading experiments under eight conditions, where the inverted load time–history curves accurately reflected the periodic characteristics of the applied loads, achieving a maximum Mean Absolute Relative Error (MARE) of 6.91%, demonstrating high stability and accuracy. The further application of the method to an in-service jacket platform in a marine environment yielded inverted wind and wave loads with a maximum MARE of 11.63% compared to loads calculated from measured wind and wave data, validating the method’s practical applicability and robustness. This approach offers a more accurate load basis for the safety assessment and residual life prediction of jacket platform structures.