arXiv subjects

Mohamad Ali Bijarchi

Publications and source records attributed to Mohamad Ali Bijarchi.

9 recordsLinked to original sources

arXiv2026

Toward accurate RUL and SoH estimation using reinforced graph-based physics-informed neural networks enhanced with dynamic weights

Accurate estimation of Remaining Useful Life (RUL) and State of Health (SoH) is essential for reliable Prognostics and Health Management (PHM), supporting timely maintenance and dependable industrial operation. However, hybrid models that combine data-driven learning with physics-based regularization often rely on fixed loss weights and therefore lose accuracy when transferred across assets with different degradation behaviors. This study introduces Reinforced Graph-based Physics-informed Networks with Dynamic Weighting (RGPD), a unified framework for spatio-temporal degradation modeling and adaptive physics-guided regularization. Graph-based representation learning captures inter-sensor degradation structure, a Soft Actor-Critic (SAC) module refines latent features under noisy conditions, and a lightweight Q-learning policy adaptively balances monotonicity, smoothness, and latent-dynamics residual losses during training. The framework is evaluated on the C-MAPSS, PHM2012, and XJTU datasets, which represent engine, bearing, and battery degradation processes. Relative to the strongest compared baselines reported in the corresponding benchmark tables, RGPD improves average RMSE by up to 12 percent on PHM2012 and C-MAPSS, and reduces average MAPE by 20 percent on XJTU compared with the second-best reported model. Performance on these heterogeneous benchmarks further suggests the model's generalizability across degradation systems. The physics-informed component is implemented through degradation-consistent priors together with a Deep Hidden Physics Model-style residual, which improves physical plausibility without requiring a full first-principles model for each asset type.

cs.LG↗

arXiv2025

An Adaptive Physics-Driven Deep Learning Framework for a Two-Phase Stefan Problem

Thermal Energy Storage (TES) using Phase Change Materials (PCMs) represents a critical technology for sustainable energy management and grid stability. This study presents a novel Physics-Driven Deep Learning (PDDL) framework for modeling the complex solid-liquid phase transition in a two-dimensional PCM-based TES system integrated with finned heat exchangers. The system operates under transient forced convection with cooling air, presenting a challenging Moving Boundary Problem (MBP) characterized by intricate phase interface dynamics and strong geometrical dependencies. Conventional numerical methods for such Stefan problems face significant computational burdens due to repeated meshing requirements at the evolving interface. To overcome these limitations, we develop a multi-network PDDL approach that simultaneously predicts the solid phase temperature field, fin temperature distribution, and the moving phase boundary position. The architecture employs three specialized deep neural networks operating in parallel, constrained by the governing physical laws of energy conservation and interface conditions. Comprehensive validation against established analytical benchmarks demonstrates the framework's exceptional accuracy in predicting interface evolution and temperature distributions across various aspect ratios. The model successfully captures the parametric influence of geometrical parameter on solidification rates and thermal performance without requiring mesh regeneration. Our approach provides an efficient computational paradigm for optimizing PCM based TES systems and offers extensibility to three-dimensional configurations and multi-material composites, presenting significant potential for advanced thermal energy system design.

math-ph↗

arXiv2025

Intelligent Optimization of Multi-Parameter Micromixers Using a Scientific Machine Learning Framework

Multidimensional optimization has consistently been a critical challenge in engineering. However, traditional simulation-based optimization methods have long been plagued by significant limitations: they are typically capable of optimizing only a single problem at a time and require substantial computational time for meshing and numerical simulation. This paper introduces a novel framework leveraging cutting-edge Scientific Machine Learning (Sci-ML) methodologies to overcome these inherent drawbacks of conventional approaches. The proposed method provides instantaneous solutions to a spectrum of complex, multidimensional optimization problems. A micromixer case study is employed to demonstrate this methodology. An agent, operating on a Deep Reinforcement Learning (DRL) architecture, serves as the optimizer to explore the relationships between key problem parameters. This optimizer interacts with an environment constituted by a parametric Physics-Informed Neural Network (PINN), which responds to the agent's actions at a significantly higher speed than traditional numerical methods. The agent's objective, conditioned on the Schmidt number is to discover the optimal geometric and physical parameters that maximize the micromixer's efficiency. After training the agent across a wide range of Schmidt numbers, we analyzed the resulting optimal designs. Across this entire spectrum, the achieved efficiency was consistently greater than the baseline, normalized value. The maximum efficiency occurred at a Schmidt number of 13.3, demonstrating an improvement of approximately 32%. Finally, a comparative analysis with a Genetic Algorithm was conducted under equivalent conditions to underscore the advantages of the proposed method.

cs.LG↗

arXiv2025

Equation Discovery, Parametric Simulation, and Optimization Using the Physics-Informed Neural Network (PINN) Method for the Heat Conduction Problem

In this study, the capabilities of the Physics-Informed Neural Network (PINN) method are investigated for three major tasks: modeling, simulation, and optimization in the context of the heat conduction problem. In the modeling phase, the governing equation of heat transfer by conduction is reconstructed through equation discovery using fractional-order derivatives, enabling the identification of the fractional derivative order that best describes the physical behavior. In the simulation phase, the thermal conductivity is treated as a physical parameter, and a parametric simulation is performed to analyze its influence on the temperature field. In the optimization phase, the focus is placed on the inverse problem, where the goal is to infer unknown physical properties from observed data. The effectiveness of the PINN approach is evaluated across these three fundamental engineering problem types and compared against conventional numerical methods. The results demonstrate that although PINNs may not yet outperform traditional numerical solvers in terms of speed and accuracy for forward problems, they offer a powerful and flexible framework for parametric simulation, optimization, and equation discovery, making them highly valuable for inverse and data-driven modeling applications.

physics.comp-ph↗

arXiv2025

Deformation and breakup of a ferrofluid compound droplet migrating in a microchannel under a magnetic field: A phase-field-based multiple-relaxation time lattice Boltzmann study

Though ubiquitous in many engineering applications, including drug delivery, the compound droplet hydrodynamics in confined geometries have been barely surveyed. For the first time, this study thoroughly investigates the hydrodynamics of a ferrofluid compound droplet (FCD) during its migration in a microchannel under the presence of a pressure-driven flow and a uniform external magnetic field (UEMF) to manipulate its morphology and retard its breakup. Finite difference and phase-field multiple-relaxation time lattice Boltzmann approaches are coupled to determine the magnetic field and ternary flow system, respectively. Firstly, the influence of the magnetic Bond number (Bo) on the FCD morphology is explored depending on whether the core or shell is ferrofluid when the UEMF is applied along α=0° and α=90° relative to the fluid flow. It is ascertained that imposing the UEMF at α=0° when the shell is ferrofluid can postpone the breakup. Intriguingly, when the core is ferrofluid, strengthening the UEMF enlarges the shell deformation. Afterward, the effects of the Capillary number (Ca), density ratio, viscosity ratio, radius ratio, and surface tension coefficients are scrutinized on the FCD deformation and breakup. The results indicate that augmenting the core-to-shell viscosity and density ratios accelerates the breakup process. Additionally, surface tension between the core and shell suppresses the core deformation. Moreover, increasing the Ca number intensifies the viscous drag force exerted on the shell, flattening its rear side, which causes a triangular-like configuration. Ultimately, by varying Bo and Ca numbers, five distinguished regimes are observed, whose regime map is established.

physics.flu-dyn↗

arXiv2025

FlexPINN: Modeling Fluid Dynamics and Mass Transfer in 3D Micromixer Geometries Using a Flexible Physics-Informed Neural Network

In this study, fluid flow and concentration distribution inside a 3D T-shaped micromixer with various fin shapes and configurations are investigated using a Flexible Physics-Informed Neural Network (FlexPINN), which includes modifications over the vanilla PINN architecture. Three types of fins (rectangular, elliptical, and triangular) are considered to evaluate the influence of fin geometry, along with four different fin configurations inside the 3D channel to examine the effect of placement. The simulations are conducted at four Reynolds numbers: 5, 20, 40, and 80, in both single-unit (four fins) and double-unit (eight fins) configurations. The goal is to assess pressure drop coefficient, mixing index, and mixing efficiency using the FlexPINN method. Given the challenges in simulating 3D problems with standard PINN, several improvements are introduced. The governing equations are injected into the network as first-order, dimensionless derivatives to enhance accuracy. Transfer learning is used to reduce computational cost, and adaptive loss weighting is applied to improve convergence compared to the vanilla PINN approach. These modifications enable a consistent and flexible architecture that can be used across numerous tested cases. Using the proposed FlexPINN method, the pressure drop coefficient and mixing index are predicted with maximum errors of 3.25% and 2.86%, respectively, compared to Computational Fluid Dynamics (CFD) results. Among all the tested cases, the rectangular fin with configuration C in the double-unit setup at Reynolds number 40 shows the highest mixing efficiency, reaching a value of 1.63. The FlexPINN framework demonstrates strong capabilities in simulating fluid flow and species transport in complex 3D geometries.

physics.flu-dyn↗

arXiv2024

Novel Magnetic Actuation Strategies for Precise Ferrofluid Marble Manipulation in Magnetic Digital Microfluidics: Position Control and Applications

Precise manipulation of liquid marbles has significant potential in various applications such as lab-on-a-chip systems, drug delivery, and biotechnology and has been a challenge for researchers. Ferrofluid marble (FM) is a marble with a ferrofluid core that can easily be manipulated by a magnetic field. Although FMs have great potential for accurate positioning and manipulation, these marbles have not been precisely controlled in magnetic digital microfluidics, so far. In this study for the first time, a novel method of magnetic actuation is proposed using a pair of Helmholtz coils and permanent magnets. The governing equations for controlling the FM position are investigated, and it is shown that there are three different strategies for adjusting the applied magnetic force. Then, experiments are conducted to demonstrate the capability of the proposed method. To this aim, different magnetic setups are proposed for manipulating FMs. These setups are compared in terms of energy consumption and tracking ability across various frequencies. The study showcases several applications of precise FM position control, including controllable reciprocal positioning, simultaneous position control of two FMs, the transport of non-magnetic liquid marbles using the FMs, and sample extraction method from the liquid core of the FM.

cs.RO↗

arXiv2024

Revising the Structure of Recurrent Neural Networks to Eliminate Numerical Derivatives in Forming Physics Informed Loss Terms with Respect to Time

Solving unsteady partial differential equations (PDEs) using recurrent neural networks (RNNs) typically requires numerical derivatives between each block of the RNN to form the physics informed loss function. However, this introduces the complexities of numerical derivatives into the training process of these models. In this study, we propose modifying the structure of the traditional RNN to enable the prediction of each block over a time interval, making it possible to calculate the derivative of the output with respect to time using the backpropagation algorithm. To achieve this, the time intervals of these blocks are overlapped, defining a mutual loss function between them. Additionally, the employment of conditional hidden states enables us to achieve a unique solution for each block. The forget factor is utilized to control the influence of the conditional hidden state on the prediction of the subsequent block. This new model, termed the Mutual Interval RNN (MI-RNN), is applied to solve three different benchmarks: the Burgers equation, unsteady heat conduction in an irregular domain, and the Green vortex problem. Our results demonstrate that MI-RNN can find the exact solution more accurately compared to existing RNN models. For instance, in the second problem, MI-RNN achieved one order of magnitude less relative error compared to the RNN model with numerical derivatives.

cs.LG↗

arXiv2024

Enhancing Convergence Speed with Feature-Enforcing Physics-Informed Neural Networks: Utilizing Boundary Conditions as Prior Knowledge for Faster Convergence

This study introduces an accelerated training method for Vanilla Physics-Informed-Neural-Networks (PINN) addressing three factors that imbalance the loss function: initial weight state of a neural network, domain to boundary points ratio, and loss weighting factor. We propose a novel two-stage training method. During the initial stage, we create a unique loss function using a subset of boundary conditions and partial differential equation terms. Furthermore, we introduce preprocessing procedures that aim to decrease the variance during initialization and choose domain points according to the initial weight state of various neural networks. The second phase resembles Vanilla-PINN training, but a portion of the random weights are substituted with weights from the first phase. This implies that the neural network's structure is designed to prioritize the boundary conditions, subsequently affecting the overall convergence. Three benchmarks are utilized: two-dimensional flow over a cylinder, an inverse problem of inlet velocity determination, and the Burger equation. It is found that incorporating weights generated in the first training phase into the structure of a neural network neutralizes the effects of imbalance factors. For instance, in the first benchmark, as a result of our process, the second phase of training is balanced across a wide range of ratios and is not affected by the initial state of weights, while the Vanilla-PINN failed to converge in most cases. Lastly, the initial training process not only eliminates the need for hyperparameter tuning to balance the loss function, but it also outperforms the Vanilla-PINN in terms of speed.

cs.LG↗