Investigating silver and alumina nanoparticles' impact on fluid behavior over porous stretching surface

The parabolic trough solar collector is among the most beneficial solar energy production technologies. However, it has comparatively low thermal performance, which can be enhanced with innovative coolant hybrid nanofluids and transmitter tube configuration. In the present investigation, water-based silver and alumina nanoparticles are used to optimize heat transfer in dual-phase flow comprising magnetohydrodynamic Prandtl-Eyring dusty nanofluid under solar radiation by employing the Levenberg-Marquardt technique with back-propagated neural networks (LM-BPNN). By combining the Joule heating phenomenon, viscous dissipation, and heat source in nanofluid, the suggested LM-BPNNs propose to enhance heat transfer. After obtaining the dataset using a numerical method called bvp4c, the Levenberg-Marquardt technique with back-propagated artificial neural networks (LM-BPANN) algorithm is employed. Benchmark datasets are used with the LM-BPANNs methodology; 80% of the dataset is utilized for training and 10% is retained for testing and verification. The generated LM-BPANNs' accuracy and convergence are verified employing the reliability obtained through effective fitness determined by mean squared error (MSE), thorough regression analysis, and suitable error histogram representations of data. With reduced MSE values of 4.38 × 10^-9, it demonstrated exceptionally good performance and demonstrated the great reliability of the model's predictions. The result demonstrates the efficacy of the suggested method and is consistent with producing a low absolute error of around zero. The main conclusions of this study should have a big impact on industries that use heat transmission, such as oil recovery, fluidic cells, solar collectors, and other related fields.