Keywords: RC buildings; structural optimization; seismic upgrading; Artificial Intelligence; Genetic Algorithm

Nowadays governments are encouraging the upgrading and the reuse (rather than the demolition) of older structures in order to reduce land use and environmental impact caused by the construction of new buildings.  The choice of the most suitable intervention for the seismic upgrading of existing structures could also be addressed combining member-level and structural-level techniques, although it may result in a complex technical challenge for engineers, since theoretically infinite combinations of various upgrading solutions could accomplish the desired structural performance. In order to support the intervention choice, an "objective" approach could be implemented making use of recently-invented Artificial Intelligence (AI) procedures. Specifically, the application of Genetic Algorithms (GAs) is usually thought as a suitable optimization procedure in several civil engineering problems. By means of a Genetic Algorithm (GA), the design of upgrading interventions results to be based on one objective criterion related to the cost-effectiveness of the intervention, rather than on the highly subjective "engineering-judgement". The goal of the present paper is to highlight the potentialities of a Soft-Computing (SC) procedure based on a GA, firstly formulated by Falcone (2017) and recently modified, which is able to increase at each iteration of the procedure the cost-effectiveness of the technically-admissible solutions that include both member-level (namely, the confinement of single columns using FRP layers) and structural-level (mainly, the realization of a concentric steel bracing system) interventions. To this aim a parametric study on a RC structure based on a similar GA procedure is reported, varying some "engineering parameters" related to the target seismic risk class of the upgraded structure. Relevant differences in the results are observed due to the variation of target of the analyses.