Automation and Optimization of Mooring Line Designs

This paper presents the study of mathematical optimization applied to the design and analysis of mooring lines for offshore structures. Mooring line design and analysis for offshore structures is carried out in either time-domain or frequency-domain. Global Maritime has developed GMoor, a commercially available frequency-domain software commonly used in industry. While GMoor streamlines the analysis, the process still requires a manual approach to iterative design; an approach that often results in designs that satisfy requirements but may not be optimal. The automation and optimization of this design problem allows for optimal solutions while saving on computational time and resources. The optimization of mooring lines is a multi-objective, multi-variable, non-linear problem. Utilizing the in-built API interface in GMoor, this study uses the ‘Constrained Optimization BY Linear Approximation’ (COBYLA) model within the NLOPT library to address this optimization. COBYLA is a gradient free optimization algorithm that iteratively approximates constraints and objective functions using linear models, making it well-suited for problems where derivatives are difficult to compute. This study sets the objective function as a minimization of total mooring line length of a semi-submersible drilling platform; the constraints are applied in line with the requirements of codes and standards for mooring line design under environmental loading, including single line failure. The optimization is first tested on a 3-mooring line design, with the optimal result validated against an exhaustive search of the design space. The algorithm is then applied to an 8-mooring line design.