Multi-objective optimisation designs for thin-walled deployable composite hinges using surrogate models and Genetic Algorithms

Thin-walled deployable composite hinges (DCHs) can achieve foldable and deployable functions by storing and releasing strain energy, which have great application potential in deployable structures, such as satellite antennas and solar wings. This paper presented multi-objective optimisation designs for DCHs. Firstly, an optimisation problem was established to obtain three conflicting objectives, minimising the peak folding moment, maximising the peak torsional moment and minimising the mass. Three design variables and one constraint had been considered. Moreover, four surrogate models were employed, including response surface methodology (RSM) and machine learning models. Root mean square error (RMSE), mean absolute error (MAE) and correlation coefficient (R²) were used to determine the surrogate model with the highest accuracy. Furthermore, four state-of-the-art Genetic Algorithms were benchmarked to obtain the optimal designs. The mimicked inverted generational distance (mIGD) was applied to determine the best optimiser. The research results have significance to practical engineering application of DCHs.