@misc {pub4166,
	title = {Surrogate-assisted Topology Optimization of Mechanical Structures},
	author = {Elena Raponi AND Mariusz Bujny AND Markus Olhofer AND Simonetta Boria AND Fabian Duddeck},
	year = {2019},
	month = {November},
	abstract = {My current research deals with the Topology Optimization (TO) of mechanical structures subjected to static and dynamic crash loads, by means of surrogate modeling techniques and evolutionary computation. TO addresses the task of determining optimal concept structures through changing the material distribution in a given design domain. Although it represents an important tool in the design and analysis of mechanical structures, TO needs much more investigation in the context of continuous global optimization of expensive and multimodal problems, e.g. associated with vehicle crashworthiness, where gradient-based methods cannot be used due to the nonlinearities, numerical noise, and discontinuities of the objective functions to be optimized. 

As such, the Kriging-Assisted Level Set Method (KG-LSM) for TO was introduced in our previous works [1,2]. It is based on an adaptive optimization strategy using the Kriging surrogate model and a modified version of the Expected Improvement (EI) as the update criterion, which allows for embedding problem-specific constraints. 

When compared to the state-of-the-art Covariance Matrix Adaptation Evolution Strategy (CMA-ES), the KG-LSM demonstrates to be significantly more efficient in terms of the number of costly evaluations, especially in the initial phase of the optimization, but with poorer capabilities of exploiting promising regions of the design space. Therefore, a hybrid technique (HKG-LSM) [3], coupling the Kriging-based optimization and CMA-ES, is currently under investigation.},
	publisher = {European Cooperation in Science and Technology},
	booktitle = {European Cooperation in Science and Technology (COST) Action Training School 2019},
	city = {Coimbra, Portugal}
}