The geometry of engineering systems affects their performances. For this reason, the geometry of systems needs to be optimised in the initial design stage. However, naval architecture and ocean engineering (NAOE) system optimisation problems are multi...
The geometry of engineering systems affects their performances. For this reason, the geometry of systems needs to be optimised in the initial design stage. However, naval architecture and ocean engineering (NAOE) system optimisation problems are multi-objective and performance analysis using commercial code is very time-consuming. Therefore, optimisation becomes difficult/impossible because we need to check the system performances for a large number of alternative design cases. To overcome these problems, many researchers perform system optimisation using the approximation model (response surface). The response surface method (RSM) is typically used to predict the system performance in the engineering research field, but it presents prediction errors for highly nonlinear systems. To create an appropriate response surface for marine systems, this thesis first compares the prediction accuracy of the response surface generated by the RSM, kriging method and artificial neural network (ANN) using a nonlinear mathematical function (user defined function and nonlinear programming) problems.
The major objective of this thesis is to establish an optimal design method for multi-objective problems in marine systems. The proposed method is composed of three parts:
- Definition of geometry
- Generation of response surface
- Optimisation process
To reduce the time for performance analysis and minimise the prediction errors, the approximation model is generated using the backpropagation artificial neural network (BPANN) which is considered as neuro-response surface method (NRSM). Then, optimisation is performed for the generated response surface using the non-dominated sorting genetic algorithm-II (NSGA-II).
Case studies of marine systems (the substructure of a floating offshore wind turbine considering hydrodynamic performance, hull form optimisation considering hydrodynamic performance, and bulk carrier bottom stiffened panels considering structure performance) verify that the proposed method is applicable to multi-objective side constraint optimisation problems in marine systems.
Evaluations of the appropriate approximation model for each problem and their application to various optimisation problems will be conducted in future work.
Keywords: Multi-objective optimisation, Approximation model, non-dominated sorting genetic algorithm-II (NSGA-II), backpropagation artificial neural network (BPANN), Neuro-response surface method (NRSM)