Multidisciplinary design optimization of dynamic positioning system for semi-submersible platform

Abstract

The dynamic positioning system (DPS) is a complex mechatronic system consisting of multiple sub-disciplines. For such highly coupled sub-disciplines and sub-systems within the DPS, the conventional sequential design strategy cannot manage this interaction adequately and explicitly. Achieving the global optimal solution becomes challenging due to the discoordination between the structure and dynamic performance parameters. To address this challenge, a co-design strategy is explored by combining the structure and dynamic performance parameters based on a simultaneous dynamic optimization approach. To minimize the operating cost of the DPS, the radial basis function (RBF) model is extended to the constraint condition. The alpine skiing optimization (ASO) algorithm is applied to solve the multidisciplinary design optimization problem. The results indicate that the co-design strategy not only reduce the operating cost of the DPS but also achieve a smaller rolling response and yaw response.