Adaptive metamodel based efficient robust design optimization of offshore structure under stochastic wave loading

Abstract

The present study deals with Robust Design Optimization (RDO) of an offshore steel structure under stochastic wave loading considering parameter uncertainty. An adaptive Moving Least Squares Method (MLSM) based metamodeling strategy has been adopted in place of direct MCS to evade extensive computational time requirement. The optimization problem is formulated to minimize the weight of an offshore deck supporting structure under maximum lateral displacement constraint. The random wave load is modelled by two approaches i.e.1) a spectral representation method with stochastic weighted amplitude wave superposition technique and 2) a simplified approach considering uncertainty in load related parameters. The RDO is formulated by simultaneously optimizing the expected value and variance of the performance function. The robustness in constraint is ensured by adding suitable penalty term to attain a specified target reliability index. The results indicate that by compromising a small increment in structural weight designer can have robust and reliable design solution within affordable computational time by the proposed RDO approach.