Artificial intelligence-aided generative design of non-imaging secondary reflector for linear Fresnel concentrating collector

In this article, an AI-aided generative design workflow is presented for the geometrical optimisation of secondary reflectors of linear Fresnel concentrating collectors. The geometry of the secondary reflector is based on principles of non-imaging optics, for which the exact analytical solution is derived for a specific linear Fresnel system acting as a validation test case. The artificial intelligence-aided generative design workflow is then applied on the same system and the optimal geometry generated is compared against the analytical solution. It is shown that the secondary reflector geometry obtained from the artificial intelligence-aided generative design workflow approaches the shape of the one obtained using the analytical solution and can outperform the latter with respect to the magnitude of radiative power intercepted by the receiver and radiative flux distribution at the receiver aperture. • An AI-based generative design workflow for the design optimization of optical devices is presented • The workflow is applied for the optimization of the secondary reflector of a linear Fresnel system • The obtained design through the workflow is compared to the one derived via analytical solution • The design obtained via the AI-workflow outperforms the design obtained via the analytical solution