The paper examines the optimisation of a three-dimensional structure formed by GL32h glulam double regular tapered beams and purlins using genetic algorithms. A software was developed in Matlab to perform the structural calculation, and an artificial intelligence optimisation tool based on genetic algorithms was tailored to optimise the roof based on cost. Finally, a statistical analysis of more than 200 optimisation data points was carried out, and several reliable prediction equations for both the geometry and the spacing of the structural elements were proposed based on some known values (i.e., span and snow load). It was found that the optimal purlin spacing is close to the maximum permissible, the optimal height and width for the purlins are 160-200 mm and 90 mm, respectively, and the optimal beam spacing falls within the range of 4-6 m.

Comunicación presentada en el XII Congreso Ibérico de Agroingeniería celebrado en Sevilla entre el 4 y el 6 de septiembre de 2023.

notPeerReviewed