Optimization of solar energy in HVAC framework using radiative nanoparticles and Levenberg–Marquardt algorithm

A comprehensive computational and artificial intelligence framework is developed in this dissertation to estimate the intricate motion and energy transmission properties of rotating Sutterby magneto-hybrid nanofluids for utilization in solar-powered heating, ventilation, and air conditioning (HVAC) system. The study examines hybrid nanofluid formulations, such as Cu–TiO2/EG mixtures, under the combined impacts of Soret–Dufour phenomena, heat radiation, Brownian motion, magnetized consequences, and permeability. The bvp4c solver in MATLAB is used to solve the model system of equations. To accurately anticipate fluid flow and thermal fields in nonlinear settings, a machine learning approach based on backpropagation neural networks (BPNN) trained via the Levenberg–Marquardt algorithm uses numerical findings as reference datasets. The ANN–numerical hybrid technique shows remarkable capacity for prediction, exhibiting substantial regression coefficients, gradients close to 10−09, and mean squared errors between 10−08–10−06. A strong tool for developing, controlling, and modifying high-performance renewable heating and cooling systems is provided by this ANN-assisted computational framework.