Influences of intelligent predictive networks on thermal efficiency of radiative magnetohydrodynamics hybrid nanofluid flow considering heat absorption-generation aspects

The research on HNFs is on peak and results revealed the efficient heat transfer properties of these liquids which are working as primary agents for industrial use. One of basic factor in improving the heat transfer of nanofluids is the mass flow rates, particle volume fractions and thermal conductivity of the nanoparticles. The improvement in thermal transmission of depending specially on the thermal conductivity of nanoparticles while flow rates and particle volume fractions remain constant. In current scenario the traditional resources of energy are found insufficient, besides this issue these resources of energy are major source of pollution on our globe. The researchers recommended that renewable energy especially solar energy is a best ultimate solution of this problem. Therefore, the goal of recent article is to use ANNs to assess the flow behavior of a Zn , TiO 2 / H 2 O hybrid nanofluid across an inclined shrinking/contacting surface. By using similarity variables, the flow models of PDEs are transformed into ODEs. Since BPLMS is used to study the graphical effects of regulating parameters on fluid velocity f ′ η and temperature distribution θ η , including the magnetic field parameter M , mixed convection parameter λ , suction parameter S , Prandtl number Pr and temperature ratio parameter . The numerical solution of HNF will be found through the usage of “ND-Solve” function in Mathematica program and produce the matrix data. By applying NN-BPLMS the graphical results are obtained in MATLAB software with the support of the matrix data. With ANNs the total samples (100 %) have been divided into the training data (70 %) and, the testing and validation (15 %) of training and validation data. By using artificial neural network on HNF, the following is the graphical outputs of MSE, error histogram, STD, regressions analysis, FSF, solution graph and error analysis outputs. The profile will be rejected as the values of suction parameter and magnetic field parameter increases. In addition, a profile grows with the values of mixed convection parameter increasing. The profile is reduced by the rising values of . Furthermore, there were increasing tendencies with increment in values of both and magnetic field parameter. • Integration of ANN-BPLMS with numerical models for HNF flow. • Up to 27 % enhancement in heat transfer with temperature ratio parameter. • Suction parameter reduced velocity profile by ~19 %. • MSE performance reaches 2.54 × 10 −11 under optimal scenario (S6). • Six scenarios reveal dominant influence of Lorentz and mixed convection forces. • ANN-trained model achieved REG = 1.000 with minimal error (≤10 −6 ).