HYDELIFE: Hybrid life cycle optimization of bridges and mixed and modular structures with high social and environmental efficiency under restrictive budgets
Principal researcher:
VÍCTOR YEPES PIQUERAS
Team members:
- JOSÉ V. MARTÍ ALBIÑANA
- JULIÁN ALCALÁ GONZÁLEZ
- MOACIR KRIPKA
- JOSÉ A. GARCÍA CONEJEROS
- IVÁN A. NEGRÍN DÍAZ
Funding Agency:
Duration: 01/09/2021 – 31/08/2024
Reference: PID2020-117056RB-I00
Summary
The economic sustainability and social development of most countries depends, among other things, on the reliable and durable performance of their infrastructure. HYDELIFE addresses the challenge of social and environmental sustainability of structures throughout their life cycle. To this end, it proposes an emerging hybrid methodology between Deep Learning (DL) from artificial intelligence, surrogate models and metaheuristics of multi-objective optimization and multi-criteria decision-making techniques. The focus applies to robust and resilient design to modular industrialized construction, both in building and in steel-concrete composite bridges and hybrid steel structures. The great challenge will be to have infrastructures capable of maximizing their social benefit without compromising their sustainability.
On the other hand, the aging of the infrastructures, the higher demand in their performance or the natural risks affect the expected performance of these infrastructures. If we add to this the deep financial and health crisis that has affected the economy of our country, the deterioration of infrastructure can cause a timely social alarm. The starting hypothesis is that the emerging hybrid metaheuristics are capable of extracting non-trivial information from the immense databases coming from the optimization and improving the quality and computation time both in the automatic design and in the optimal maintenance of bridges andstructures. This methodological proposal aims to address the uncertainties of the real world by proposing optimal design and maintenance based on reliability and robust designs. This hypothesis should be extended to multi-criteria decision-making processes that address social and environmental sustainability for the entire life cycle that take into account fluctuations in both parameters and possible scenarios, especially in the case of severe budgetary restrictions. This methodology presents, however, serious difficulties, so surrogate models and DL capable of accelerating complex calculation processes must be explored. Furthermore, it is intended to deepen in the emerging multicriteria decision techniques such as neurosophic logic and others such as Bayesian networks. In this context, although progress has been made in multi-objective optimization of structures, in the real world, there are uncertainties, imperfections or deviations from the parameters used in the codes. An optimal structure is close to the infeasible region, so it is necessary to incorporate the uncertainties to provide more robust and reliable designs.
On the other hand, the severe budgetary constraints, which are present in moments of economic recession deeply, compromise the infrastructure generation and maintenance policies. For this reason, after a sensibility analysis of different budgetary policies, the results of this project aim to determine which typologies, specific actions of maintenance and demolition and re-usable alternatives are suitable for minimizing environmental and social impacts taking into account the variability. In this respect, an important aspect is to determine which criteria and indicators are key to include, in an effective manner, the sustainability in the procurement procedures of new projects and maintenance of modular buildings and steel-concrete composite bridges.
