Novel Machine Predictive Exogenous Knowledge Driven Neuro-Structures for Unsteady Squeezing Nanofluidic Model with Rotating-Oscillating Disks

Abstract

Artificial intelligence plays a significant role in demonstrating nanofluidic systems through analysis of the large datasets for data-driven insights, improving prediction accuracy through iterative learning, aiding in design optimization, and the development of nanofluidic devices with superior thermal radiation heat transfer characteristics. This study investigates heat transport in the flow of unsteady squeezing nanofluidic model with stretchable rotating and oscillating disks mixed with kerosine oil as a base fluid by using artificial intelligence-based knacks through nonlinear autoregressive networks with Levenberg–Marquardt backpropagation. The partial differential equations are converted into ordinary types by changing multi class parameters, i.e., stretching, squeezing, and rotation, with fixed numbers, i.e., Hartmann, Eckert and Prandtl. The synthetic dataset is generated with Adams numerical method for unsteady squeezing flow and heat transport of Silicon oxide nanofluidic model and further this information is utilized for the execution of nonlinear exogenous networks for solving the unsteady squeezing nanofluidic model. The results are consistently aligned with numerical solutions for the system, demonstrating a substantially reduced error magnitude across several anticipated scenarios. The effectiveness of the proposed methodology is demonstrated through iterative convergence on mean square error, adaptive controlling metric of optimization with Levenberg–Marquardt algorithm, statistical distribution of error in histogram plots, and autocorrelation analysis on exhaustive numerical experimentation of the nanofluidic model.