PROJECT TYPE:

• Research project

PROJECT DURATION (start and end date):

• 1.10.2025. – 30.9.2029.

PROJECT LEADER AND COLLABORATORS:

Project Leader:

• doc. dr. sc. Damir Karabaić

COLLABORATORS:

• prof. dr. sc. Sven Maričić, Tehnički fakultet, Sveučilište Jurja Dobrile u Puli, Medicinski Fakultet, Sveučilište u Rijeci
• izv. prof. dr. sc. Marko Kršulja, Tehnički fakultet, Sveučilište Jurja Dobrile u Puli
• izv. prof. dr. sc. Lovro Liverić, Tehnički fakultet, Sveučilište Jurja Dobrile u Puli
• naslovni asistent Matija Bauer, mag. ing. mech., Tehnički fakultet, Sveučilište Jurja Dobrile u Puli
• naslovni asistent Miralem Mešanović, mag. ing. mech., Tehnički fakultet, Sveučilište Jurja Dobrile u Puli

PROJECT SUMMARY

Modern engineering simulations, although crucial for product development and process optimization, often require significant execution time and extensive computational resources. This project will focus on innovative approaches that combine advanced optimization techniques and machine learning (AI) models to address this issue. Over four years, researchers will develop, implement, and validate new methodologies within the framework of commercial software such as ANSYS OptiSLang.

These methodologies will be applied to a wide spectrum of engineering problems, including parametric optimization, robustness analysis, model dimensionality reduction, and the creation of surrogate models. The project is expected to result in a significant reduction in simulation costs, a shortening of development cycles, and an increase in industrial innovation.

PROJECT OBJECTIVES

The main goal of the project is to develop and validate improved methodologies for applying optimization and artificial intelligence (AI) in engineering simulations. The focus is on reducing the required number of simulations, the individual time for each simulation, and the total simulation execution time, while simultaneously maintaining high accuracy and reliability of the results. Specific goals include:

1. Validation and enhancement of existing integrated optimization methodologies and the development of new variants, including the requirements of multi-objective optimization (optimization with multiple conflicting goals).

2. Improving the efficiency of optimization algorithms: Research and adaptation of advanced algorithms (e.g., metaheuristics, surrogate model-based algorithms, machine learning) for specific engineering problems, with the aim of reducing the number of required simulations and the overall execution time.

3. Validation and application of the developed methodologies to relevant engineering case studies from various disciplines (e.g., mechanical engineering, aerospace, biomedicine) for the purpose of validating and demonstrating their effectiveness.

​​​​​​​

PROJECT HOLDER: Juraj Dobrila University of Pula