Structural Health Monitoring of Subsea Pipelines by Genetic Algorithms

The measured dynamic response can be used to calibrate and update the numerical model used in the design so as to better predict the structural response and assess the structural safety, by means of a suitable system identification strategy, which can be further applied to monitor the “health” of the structures continuously. Recently, structural health monitoring (SHM) of offshore structures has increasingly gained more and more attention as it can provide useful information to predict potential damages or accidents. This paper presents a numerical study on subsea pipeline bedding condition assessment, focusing on pipe integrity and free span identification. Since subsea pipelines are usually very long without clear boundaries, only a span of pipelines is modeled to study. Improved Genetic algorithms (GA) method is adopted to identify the key parameters, due to its advantages of robustness and ease of implementation. Output-only substructural identification is formulated and applied. Numerical study is conducted by Fortran programming, and it is shown that the proposed identification strategy is effective for global system identification of subsea pipelines.