An evolutionary optimisation study on offshore mooring system design

This research was undertaken to provide tools to aid in the design and optimisation of offshore mooring systems. The tools present a set of feasible solutions with different performance characteristics. With an increased reliance on the exploitation of the reserves of hydrocarbon on marginal offshore areas, both in Australia and around the world, mooring mobile vessels becomes a more challenging task. The latest evolutionary optimisation techniques offer new prospects for designing and analysing moorings, but to select an appropriate method of optimisation, both the classical deterministic search and evolutionary optimisation methods are reviewed. Genetic Algorithms (GAs) and Particle Swarm Optimisation (PSO) have been investigated in terms of their capability, strengths, and limitations. The GA and PSO have been benchmarked against each other, and PSO has been found to more efficient in the engineering applications considered. Particle swarm optimisation was selected and implemented in the evolutionary mooring design computer software coded by the author. This software tool interfaces with the offshore industry commercial package OrcaFlex, which supplies results from model simulations. A case study of a mooring design is presented to demonstrate the ability of the optimisation tool, while the results provide potential improvements in mooring design solutions. A performance based multi-objectives particle swarm optimisation (MOPSO) was developed where the multiple constraints can confidently be switched to problems with multiple objectives. Meanwhile, the results from the mooring optimisation obtained from MOPSO were compared with PSO, and showed that highly competing constraints leads the PSO search in an unhealthy manner. Alternatively, the performance based MOPSO design produces a set of solutions with a different performance which can be further processed and selected by offshore engineers. MOPSO offers greater flexibility and computational efficiency to optimise the final solutions.