arXiv subjects

Raghav Kumar

Publications and source records attributed to Raghav Kumar.

2 recordsLinked to original sources

arXiv2024

AI-Driven Physics-Informed Bio-Silicon Intelligence System: Integrating Hybrid Systems, Biocomputing, Neural Networks, and Machine Learning, for Advanced Neurotechnology

We present the Bio-Silicon Intelligence System (BSIS), an innovative hybrid platform that integrates biological neural networks with silicon-based computing. The BSIS, a Physics-Informed Hybrid Hierarchical Reinforcement Learning State Machine, employs carbon nanotube-coated electrodes to interface rat brains with computational systems, enabling high-fidelity neural interfacing and bidirectional communication through self-organizing systems in both biological and silicon forms. Our system leverages both analogue and digital AI theory, incorporating concepts from computational theory, chaos theory, dynamical systems theory, physics, and quantum mechanics. Additionally, the BSIS replicates the neuronal dynamics typical of intelligent brain tissue, employing nonlinear operations underlying learning and information storage. Neural signals are read through the FreeEEG32 board and BrainFlow software, then features are extracted and mapped to game actions by tracking feature changes in continuous data. Metadata is encoded into both analogue and digital brain stimulation signals at the microvolt level using our proprietary software and hardware. The system employs a dual signaling approach for training the rat brain, incorporating a reward solution and sound as well as human-inaudible distress sounds. This paper details the design, theory, functionality, and technical specifications of the BSIS, highlighting its interdisciplinary approach and advanced technological integration.

q-bio.NC↗

arXiv2022

Formulation of an equivalent GNF model as an efficient approximation for flow of polymer solutions described by FENE-P

The molecular constitutive models, like FENE-P for polymer solutions, are known to have convergence issues at relatively larger flow rates. In this study, we investigate the possibility of a numerically efficient GNF-based approximation of FENE-P, which would closely approximate the flow field. Firstly, we compare the flow fields predicted by FENE-P and an equivalent GNF model. For these studies, we considered the flow around a sphere and selected the Carreau-Yasuda model as the representative GNF. This is made equivalent to the FENE-P by selecting parameters to equalize the viscosity-shear rate dependence. Our results show severe deficiencies of the GNF model, owing to its inability to account for chain stretching, particularly near the stagnation points. Next, the effect of extensional components on the local viscosity was added by formulating an equivalent GNF-X [Journal of Rheology 64, 493 (2020)] model. Even this failed to capture the asymmetry in the stress and flow profiles and predicted very large stresses at both stagnation points, relative to FENE-P. Hence, we proposed a novel modified formalism (denoted as GNF-XM) that was able to capture all trends successfully. The drag coefficients from GNF-XM agreed well with FENE-P predictions for all flow rates considered. Significantly, the computational times required to solve the flow field with GNF-XM is about an order of magnitude lower than that of FENE-P, especially at higher flow rates. Thus, we have successfully formulated a highly efficient GNF-based approximation to the FENE-P, whose formalism can be extended to other similarly complicated constitutive models.

physics.flu-dyn↗